Random Thoughts

Why does CORE fail? Part 2

2010-05-03T20:10:00.000-07:00

... Continuation of my previous post on CORE deficiencies and how it could be improved upon.

What is CORE?

Let's look at originally defined CORE equation.

CORE = (S x R x V)/(C x tc)

where,

S = The capacity being reduced in TB. Dave in his post fixes the S value at 100TB to compare all solutions.

R = The percent reduction achieved. Dave shows the R value in decimal for different solutions, we can assume though R is described as percent reduction, decimal R is used in calculating CORE.

V = The value of capacity being saved. Though, Dave doesn't list the V values used for different solutions, it is not difficult to reverse-calculate this value using other parameters listed in his table.

C = The cost of solution doing the reducing.

tc = The elapsed time to compress the capacity. As covered in my last post, I consider this parameter to be stated incorrectly, incorporated inappropriately and irrelevant to the CORE. In place, a better parameter would have been the elapsed time to write.

Three things stand out in this CORE equation:

CORE equation assumes first-order relationship with its variables. It may seem that for a specified value of S, the high CORE score can be achieved by achieving high data reduction (R) and the value of capacity being saved (V) and reducing the cost of solution (C) and the time to compress (tc).

CORE equation has variables (S, R, V) in numerator that are normalized for solutions without data reduction but no such adjustment is made for variables (C, tc) in denominator.

CORE equation is composed of dependent variables instead of independent variables.

Isn't V dependent on S and R?

V is defined as cost per TB (Ct) times amount of data reduced (Sr), according to the description of the math for CORE. Amount of data reduced (Sr) is the capacity being reduced (S) times percent reduction achieved (R).

V = Sr x Ct = S x R x Ct

Substituting V in original CORE equation,

CORE = (S^2) x (R^2) x Ct / (C x tc)

To a large extent, this modified CORE equation is composed of more independent variables than original one. Obviously, it is no longer a first order relationship with S and R.

What is interesting with CORE equation is that amount of data reduced has been included twice, once as amount of data reduced and then again as part of cost of amount of data reduced.

What is the CORE value for a solution with no data reduction technology?

For,

S = 100 TB,
R = 0% as there is no data reduction,
V = $0 as there is no capacity being saved,
C = 0 as there is no data reduction technology in play so there is no cost of data reduction solution, and
tc = 0 ms as there is no compression of data taking place,

CORE = (S x R x V) / (C x tc) = (100 x 0 x 0) / (0 x 0) = 0/0

CORE = 0/0 (indeterminate) this expression has no meaning.

You may agree that a relevant CORE value for a solution with no data reduction technology should be 0 or 1. It also makes sense in calculating value of a data reduction solution to have a solution with no data reduction as baseline.

How can we avoid division by zero?

1. Replace tc with tw or (tw + tc)

An equation that takes in to account time to write (tw) instead of or in addition to time to compress (tc) could help avoid division by zero when there is no compression/deduplication being used as even baseline solution with no data reduction will have a non-zero time to write. Either tw or (tw + tc) will be a better choice in place of tc in original CORE equation.

2. Redefine tc and tw

Of course, as originally defined in Dave's post, tc is time to compress the smallest unit compressed in the solution (e.g. file or multiple files or blocks) which ignores the variation in tc due to variation in the size of smallest unit across various solution. I recommend changing the definition of tw and tc, respectively, to time to write and to compress S amount or certain % of S, the value of S should remain same across all solutions. This will remove the parameter dependency on smallest unit compressed and normalize parameter across same amount of S.

3. Redefine C

As originally defined, C is the cost of data reduction solution. As Dave' post indicate NetApp doesn’t charge for ASIS – we took a percentage of the array’s cost, we can safely assume that C is only the cost of data reduction part of the solution, and not the whole solution. In this scenario C = 0 for a solution with no data reduction, thus making CORE value indeterminate again.

An equation that takes in to account the total cost of solution, i.e. cost of solution with no data reduction plus the cost of data reduction solution will help avoid division by zero. Of course, for a data reduction solution that uses existing storage, the total cost of solution will be net present value (NPV) of existing storage plus the cost of data reduction solution. Even better, subtract cost of capacity saved (V) from this cost instead of using V in numerator will result in Net cost of solution.

A better CORE equation, may be?

CORE = (S x R) / (C x tw x tr)

where,

S, R and V are same as originally defined.

C = Net Cost of Solution = Cost of data reduction solution + Cost of capacity used after reduction

Cost of capacity used after reduction = S (1 - R) x Ct = (S x Ct) - (S x R x Ct) = (S x Ct) - V

tw = time to write a pre-defined storage capacity or fraction of S

tr = time to read a pre-defined storage capacity or fraction of S

Of course, some may object to not including read/write ratio, there is no reason why read/write ratio shouldn't be included.

In the end, a CORE equation that is function of Storage Capacity (S), Percent data reduction (R), Net Cost of Solution (C), Read/Write ratio, Time to write (tw), and Time to read (tr) will be more valuable than the originally defined CORE equation. Of course, a lot more work is required to determine the interdependency of these variables.

Why does CORE fail? Part 1 - Response

2010-04-29T21:40:00.000-07:00

Steve Kenniston of Storwize made detailed comment in response to my last post Why does CORE fail? Part 1. I thought my response to his comment deserved a separate blog post. Frankly, I haven't kept up with developments at Storwize since May 2007 when I last wrote a series of blog posts on Storewiz so I don't claim any knowledge of current Storwize solution.

First, I am not so sure that time to 'uncompress' ... is a valid parameter IF all solutions are being compared identically,....

The time to decompress/reconstitution is as much important, if not more, than time to compress/dedupe. The compression/deduplication can be managed 'internally' to keep up with write expectations of applications and users whether through delaying writes just enough to allow data reduction in-band or through data reduction after writes complete or some hybrid approach. But, the read expectations must be met in-band so any decompression/reconstitution need to take place correctly and completely in the expected time. A solution that requires lower time to decompress should be rewarded in same fashion as a solution with lower time to compress being rewarded in CORE.

... First I think we can all agree that decompression or rehydration is faster than optimization (compression, deduplication). ... the performance of time to 'compress' (I prefer optimize) and then cut the time in half and call this time to rehydrate. Now apply the formula. I would assume that the new CORE value would come out very close as they are now.

I am not so sure of time to decompress/reconstitute being faster than time to compress/dedupe or being 50% of time to compress/dedupe as I haven't heard of a solution or seen data yet that supports such claim. Actually, the relationship may be reverse specially for solutions with large amount of compressed/deduped data and high data reduction ratio. Only related published data, I am aware of, is that of read speed being direct function of the smallest unit used for decompression/reconstitution - larger the unit size, higher the read speed.

As I questioned in my last post, are time to decompress and compress proxy for time to read and write from data reduction solution? If it is the case, CORE could be improved upon by including actual time to read and write (instead of time to decompress or compress) or including time to decompress/compress as penalty over normal read/write with a solution that has no data reduction technology - in essence, additional cost in the form of lower read/write performance in exchange for higher storage efficiency.

Also, without understanding how the solution works it is very difficult to debate the merits of the value of performance on that solution. ...

If CORE stays with the parameters that can be judged externally for a solution, it will be more relevant and valuable than trying to incorporate parameters internal to a solution like time to compress (tc). A CORE based on externally measured parameters like reduction ratio, read and write performance, and cost of solution over a range of storage capacity and time may produce a better value indicator. Any attempt to include internal mechanisms weakens the CORE due to lack of complete information and understanding of every solution and rapid changes in technology and techniques incorporated in such solutions.

How can you possibly say that a post process solution that has users: 1) Buy full storage capacity (vs. less capacity with an inline solution) ...... is a good solution? ...

Please read my post again. I never claim any one solution is better than other. CORE includes cost of solution as a parameter which supposedly should penalize the solution that includes more storage than required by other solutions.

Step out of the vendor shoes for a moment and put yourself in the shoes of the customer. Which would you want?

As a customer, I want a solution that will provide additional storage efficiency at reasonable cost while meeting my expectations for read and write performance, safeguards my data and doesn't require additional management overhead. Anything beyond that is vendor coloring the customer expectations to fit it's solution.

Why does CORE fail? Part 1

2010-04-26T18:21:00.000-07:00

Recently, David Vellante at Wikibon wrote in his blog post Dedupe Rates Matter ... Just Not as Much as You Think about his Capacity Optimization Ratio Effectiveness (CORE) value for ranking dedupe/compression/capacity optimization solutions. He also applied CORE to few dedupe solutions for primary storage.

As I commented on his blog, right away I noticed that CORE formula had an important parameter missing - time to uncompress/reconstitute (hereafter referred as time to uncompress) deduped data. It is an important parameter that impacts the rate of reading data from dedupe solution by applications/users. As time to uncompress need to be happen inline for both inline and post-processing solutions, logically there will be no major discrepancy in using time to uncompress and reading data from a dedupe solution interchangeably.

Is time to compress/dedupe also proxy to rate of data written to dedupe solution?

Another important parameter is rate of writing data to a dedupe solution as applications/users have certain expectations on how quickly data must be written to a storage system. David includes time to compress (tc) in his CORE calculation, what I assume, as a proxy to rate of data written to dedupe solution. I may be wrong as I didn't see an explicit statement about why time to compress/dedupe is important.

In my opinion, he incorrectly assumes the impact of time to compress/dedupe (hereafter referred as time to compress) to be same across various dedupe solutions whether inline or post processing solutions. The time to compress impacts the rate of writing data, more so, for a dedupe solution that uses inline processing. There is no impact on rate of writing data for post-processing solutions. So, to have apple-to-apple comparison, David need to either use the rate of writing data across all solutions or include time to compress data as penalty for inline solution due to slowing down the rate of writing data.

The low time to write data is a requirement of applications/users which inline solutions meet by reducing the time to compress as much as possible (possibly at the expense of lower dedupe ratio). Post processing solutions meet the same requirement by delaying the compression/deduplication for later (possibly at the expense of additional capacity required for storing pre-deduped data).

Including time to compress data in CORE calculations without discrimination inaccurately biases the CORE toward inline solutions. Just because a solution have sub-ms time to compress in-band doesn't mean it should be rewarded over a solution with few ms time to compress out-of-band.

Assuming that time to compress in inline mode and post processing mode are equivalent, in CORE calculation, is flat out incorrect.

Why is Time to Compress being used as Time to compress the smallest unit compressed in the solution (e.g. file or multiple files or block)?

Is a dedupe solution that compresses 16KB block size in 0.001ms better than a solution that compresses 64KB block size in 0.003ms? The CORE fails right here.

For all other factors being equal, a solution that claims 0.001ms for compressing 16KB (smallest unit for the first solution) will produce higher CORE value than a solution that claims 0.003ms for compressing 64KB (smallest unit for the second solution). As specified currently, the time to compress, in turn CORE, doesn't take into consideration the variation in different unit size used by different solution. Is the CORE formula assuming that compressing/deduping in smaller units better than in larger units?

The smallest unit compressed varies across solutions by a wide range, even >1000x factor. The time to compress should be the amount of time it takes to compress a specified storage capacity and should be normalized across all solutions for CORE to be of any value. Comparing time to compress 16KB units versus 64KB units is like comparing oranges-to-apples. For 1MB data, in first case 64 units will need to be compressed (0.064ms) versus 16 units in later case (0.048ms). CORE using time to compress/dedupe without taking into consideration the unit size will penalize the second solution incorrectly.

In next post, I will further look in to CORE and take apart CORE formula ...

Spreadsheet Miscalculation of 30!

2009-07-19T14:56:00.000-07:00

Today, I encountered something amusing. I was trying to calculate Factorial 30 (also written as 30!). Don't ask me why I was trying to calculate Factorial. :-).

As a refresher for some of us who may have forgot Factorials, Factorial 30 is product of all positive integers between 1 and 30 inclusive.

30! = 30 x 29 x 28 x 27 x 26 x 25 x 24 x 23 x 22 x 21 x 20 x 19 ... x 1.

With Microsoft Excel, using both PRODUCT function and manual multiplication 30 x 29 x ....

30! = 265,252,859,812,191,000,000,000,000,000,000

Same results on my MacBook with Numbers program, using both PRODUCT function and manual multiplication.

30! = 265,252,859,812,191,000,000,000,000,000,000

I got slightly different result with Google Docs Spreadsheet, using PRODUCT function

30! = 265,252,859,812,191,030,000,000,000,000,000

And, another different result with Google Docs Spreadsheet using manual multiplication

30! = 265,252,859,812,191,100,000,000,000,000,000

But actually,

30! = 265,252,859,812,191,058,636,308,480,000,000

It appears spreadsheets are rounding numbers after 15 or 16 digits.

Peril of Working in Cloud

2009-05-31T18:44:00.000-07:00

Sorry! We are experiencing technical difficulties and cannot show all of your documents.

Do you have local backup copies of everything important you store in the Cloud?

Any Vendor Strategy, why not?

2008-12-11T20:59:00.000-08:00

Initially, I was going to post a comment on Chris Evan's recent post 2V or Not 2V (vendors this is). With the increasing length of the comment, I decided to turn it in to a blog post of my own. Chris succinctly covered the operational aspects and challenges of multi-vendor strategy.

The challenge is how deep do you go in your environment to have multiple vendors. Do you want to have multiple vendors for,

only large items like storage subsystems?
smaller stuff like HBAs and switches too?
commodity type stuff that has little differentiation among vendors?
specialized products?

Just because you have multiple vendors, doesn't necessarily gives you $ bargaining power. Bargaining power comes with the transaction volume, transaction size, transaction frequency and your value to the vendor.

At the smaller end, though you can achieve better operational efficiency by standardizing on single vendor, you don't have the volume and size for a single vendor to take you seriously. Unless by consolidating all your purchases you get the volume and size to be valuable to a vendor, why not just buy the best-of-breed solutions?

How much operational efficiency are you going to gain by buying three Clariion versus one Clariion, one 3Par and one Compellant?

At the high end, single vendor strategy hinders your ability to adopt innovation and new technologies with minimal gains in operational efficiency (remember large teams can be split among multiple vendors if needed) though you may be valuable to the vendor and get better pricing. How much operational efficiency are you going to lose by adding three 3Pars to couple of dozen AMS, you already have?

I have seen, heard and experienced enough horror stories to believe either single or multiple vendor strategy for any one organization is a right strategy. I favor Any Vendor strategy where your decisions are driven by the best solution that meets your need and not a solution from a pre-selected vendors that somewhat meets the needs.

Adaptec Advisors are Back!

2008-11-25T21:10:00.000-08:00

Adaptec PR firm sent a note mentioning that Adaptec Storage Advisor's blog is back! Check it out.

I am also trying to get back to updating my blog after a long hiatus. Hopefully with some small and quick blog posts on regular basis, my writing habit will establish. In the mean time, enjoy the sights from my various trips.

How do you overcome writing drought?

Online Backup Services - Six Questions

2008-07-14T23:23:00.000-07:00

During my visit to Denver few weeks ago, I had the opportunity to talk with folks working with online backup and archive cloud services. Some of my impressions from these discussions are interesting and worth sharing. These are based on what I heard from professionals working for or providing services to online backup service providers. These are not result of a full-blown survey, and at best anecdotal. You are welcome to respond to these questions if you like via comments, emails or your own blog post.

Q1: Who are the primary adopters of Online Backup Services?

Individuals and small businesses.
Entities with fewer than a dozen workstations .
Few with a centralized server.

Q2: What was the primary backup method before adopting online backup?

None.
A USB key or USB attached disk drive.
Few with a share on another workstation.

Q3: What was the offsite backup strategy before adopting online backup?

None.
A Floppy, USB key or CD with important files.
Few with a mobile HD.

Q4: What is the subscription and retention rates for online backup service?

High subscription rate.
Very low retention rate.
Most abandoned service within few weeks.

Q5: What are the primary reasons provided for discontinuing use of online backup service?

Excessive use of Internet connection.
Backup takes too long.
Poor experience during primary use of workstation.

Q6: What was the backup method after discontinuing online backup?

A USB attached disk drive.
A NAS device on network.
Few with no backup method.

Summary

Overall, online backup services seems to be a great way to introduce backups to people with no prior backup methods as only few reverted back to no backups after discontinuing use of online backup service. Tape is non-existent in environments that are finding online backups attractive. Despite heightened awareness of online backup service, the low bandwidth connection to Internet continues to be main hurdle in retaining subscribers, a focus on spending limited resources on sales improving or cost reducing services over a fear-based buying decision. A comment I heard was,

I prefer to allocate 50% of Internet bandwidth to VoIP services that reduce my telecommunication cost instead of to offsite backup.

Denver Visit, New Piñata & Scalability Videos

2008-06-22T09:33:00.000-07:00

Week in Denver

I will be in Denver this week till Friday June 27th. Unfortunately, I will miss nPost Golf 2.0 event in Seattle.

Despite a busy work day schedule in Denver, I am looking forward to seeing some friends and colleagues also. If you are a fellow storage blogger or reader or working on a cool storage technology and located in Denver area, ping me and we can meet one evening during my visit.

New Piñata for EMC & IBM

Recently, a reader alerted me to new Data Domain blog Dedupe Matters written by Brian Biles. Welcome Brian to the world of Bloggers. Lets see how quickly EMC and IBM bloggers make you the new piñata like they did to HDS bloggers. ;-)

In any case, it’s a nice change from their rumored no blogging policy. Hopefully, blogging at Data Domain will go beyond people in Ivory Towers.

Video of Presentations from Google Scalability Conference

Google already uploaded the videos of presentations from last week's Google Scalability Conference. I also plan to discuss some of the presentation topics in further details as time permits.

Welcome Remarks by Brian Bershad

GIGA+ by Swapnil Patil

HPC with NetworkSpaces for R by David Hendersen

Chapel by Brad Chamberlain

SMP via Transactional Memory by Vijay Menon

Communicating Like Nemo by Jennifer Wong

Maidsafe by David Irvine

CARMEN by Paul Watson

Scalable Wikipedia by Thorsten Schuett

Google Conference on Scalability - First Impression

2008-06-16T23:38:00.001-07:00

As expected from conference schedule, Google conference turned out to be a technical event primarily focused on parallel programming and infrastructure scalability. At last minute, Google decided to merge two tracks in to one. Though, I got to attend all the sessions, they felt time-compressed and rushed. I was surprised to see lot of attendees who came from outside Seattle. I met quite a few people from Bay area, Canada and Europe. I enjoyed the sessions though some audience members commented about very technical nature of the conference compared to previous year. As Brian Bershad, Google commented in his welcome speech, the challenge is to find technologies and solutions to scale handling search queries from 600 million to 6 billion. And, I came away better informed on different challenges and potential solutions we may see down the road.

I also sat down and chatted with Robin Harris. We decided to forego making a video of our conversation. I am not a big fan of talking head videos or podcasts unless they leverage the unique values of these methods not available through written words or pictures. And who wants to listen to two storage bloggers chatting about nothing. I find them miserable myself so why put others through the same misery.

In my opinion, three sessions: CARMEN: a Scalable Science Cloud [PDF], GIGA+: Scalable Directories for Shared File Systems [PDF] and maidsafe stood out at the conference from infrastructure scalability perspective. Communicating Like Nemo was very entertaining. The common theme in audience questions on most infrastructure presentations was reliability, availability, scalability, and security of the offered solution. It is a good indication of what is on the mind of people when evaluating new infrastructure offerings. With the popularity of hashing in storage of data, speeding up hash lookup is becoming an interesting problem for scalability.

David Irvine's session on maidsafe was the only session where a speaker white-boarded most of the presentation. His confidence and knowledge was commendable. Not many speakers can pull off white-boarding 80% of presentation with 100s in audience. Comparing maidsafe with ant colony was an interesting way to show scalability and simplicity of solution. Maidsafe solution seems to be in same category as RevStor, Seanodes, Cleversafe, Oceanstor, Farsite and several others that are trying to leverage storage across 100s and 1,000s of distributed nodes in a peer-to-peer or quid pro quo network, a solution most likely attractive to players in cloud and web distribution market.

Heading to Google Scalability Conference

2008-06-12T23:30:00.001-07:00

Like Robin Harris, I am also attending Google Scalability Conference in Seattle Friday and Saturday. Hopefully, I will see him at the event. All sessions at Google conference look good. Unfortunately, I will miss half of them as two tracks are running in parallel. I am looking forward to hearing about maidsafe, CARMEN, GIGA+, and Google Maps scale down.

Recently, several readers inquired whether I lost interest in blogging as I am posting very irregularly. Nothing can be further from the truth. I am still as excited about blogging as I was almost five years ago when I first delve into blogs. The lack of frequent updates is due to my attention being somewhere else (new dig, rig and gig). Will elaborate some other time.

BTW, checkout Storage Optimization blog by Carter George. His startup Ocarina got interesting story with their data footprint reduction technology. I am looking forward to learning more once I refocus on new exciting stuff in storage. Hopefully, Ocarina can change the KPCB's luck in storage space.

Do You Got Talent?

2008-05-06T22:28:00.000-07:00

After a long day of work, this clip made me laugh. Nothing to do with data storage, just a performance from British TV Show, Britain Got Talent.

Online Backup: 100% Install

2008-04-17T00:55:00.001-07:00

My last post Online Backup any different from Traditional Backup for Laptop/Desktop? was quickly turned in to us vs. them argument by Beth Pariseau in her blog post Blog dialogue: Online vs. traditional backup. I guess my curiosity and conversation starter about slow adoption of online backup didn't come across clearly.

… Gupta probably has “too much” experience with backup clients to necessarily see things from the SMB customer’s point of view. For him, installing a backup client isn’t a big deal–for some, it might be enough of a reason to let somebody else deal with it.

Initially, I thought about pulling Tony on her. On a side note, I wonder why Tony spills coffee every time Hu sneezes.

More I analyzed her statements, more I realized her opinions most likely resulted from what she heard as a storage news writer and from whom instead of her own experiences. Keywords like SMB are a good giveaway whom she is listening to. Not many practitioners try to segment customers with mile-wide brush. ;-)

Lets start with addressing her installation related concerns. Do online backup services magically appear and start working on your laptop/desktop by themselves? No, someone has to download and install them. Only backup clients that come pre-installed on your system are the ones that don't require install. As I understand, there are two main backup clients available that don't require installation and readily available to users, one provided by Microsoft with Windows XP (Windows Backup) and other one provided by Apple with Leopard (Time Machine).

Lets add configuration of the backup client to the part of "difficult to install" equation. Configuration of Mozy Pro [PDF 46 pages] and Windows Backup [Web page - 6 pages if you decide to print], are available online for your review and comparison. Of course, Time Machine is so simple to configure that even someone like me, who misunderstands backup needs of SMB according to a marketer, implemented on MacBook without instructions. BTW, AppleInsider article Road to Mac OS X Leopard: Time Machine is a good overview of Time Machine.

You be the judge how difficult each one is to install and configure.

As I wrote in my comment on Beth's blog, my intention is not to promote one method over another, just to show similarities and question the current implementations. Hopefully, these posts are setting the stage for future opinions and conversations that will help improve current BaaS offerings and develop new ones.

More to come.

Online Backup any different from Traditional Backup for Laptop/Desktop?

2008-04-10T22:11:00.001-07:00

Recently, Beth Pariseau wrote in her blog post HP unveils unlimited online storage for SOHO market that bandwidth is one of the hurdles in adoption of online backup services.

Like most online storage offerings to date, this offering is small in scale and limited in its features when compared with on-premise products. Most analysts and vendors say online storage will be limited by bandwidth constraints and security concerns to the low end of the market, with most services on the market looking a lot like HP Upline.

Though, it is a validation of my thoughts expressed in blog post Bandwidth, one hurdle in adopting Cloud Storage, I am not totally convinced of bandwidth being the root cause of limited adoption. There may be something else hindering adoption of online backup services.

Recently, Scott Waterhouse, an EMC blogger also has been discussing the virtues of Mozy, an online backup service (acquired) by EMC. I agree with his argument about the challenges of traditional backup clients in post Mozy as the Future of Backup.

Big business has a lot of data on laptops and desktops. Traditionally, installing backup clients on these systems has been costly, full of headaches, and generally causes more problems than it solves. The consequence of this is that most folks just don't protect them.

Is Mozy client any different? Is there any difference in installing, configuring, using and maintaining traditional backup client versus Mozy client on laptop/desktop? Nothing, I noticed after reading his posts.

My intention is not to pick on Mozy or Scott but there is nothing unique in most Online Backup Services that couldn't be in traditional backup for laptop/desktop. At least traditional backup also come with peace of mind that all backups are stored on company's own infrastructure. In last few years, I tried over a dozen online backup services in addition to putting up with traditional backup clients for laptop/desktop and I don't see much difference among the two.

IMO, most online backup services are just taking existing on-premise backup strategy for laptops/desktops and repackaging it to run backups to somebody else's infrastructure instead of your own. What do you think?

Storage Jobs @ Startups

2008-03-27T22:28:00.000-07:00

Recently, Nathan Kaiser at nPost contacted me regarding his new widget displaying Startup Jobs on blogs. As sidebar on my blog is already too long, I decided to include his widget in a blog post. Try it out and let me know your feedback (positive and negative).

P.S. If you are using a RSS reader like Google Reader and don't see the widget, please visit my blog. While I am writing this post, I am not sure if widget will show up in the blog post either. In case it doesn't, please visit nPost Startups Jobs site to check out the startup jobs. Use keyword "storage" to find storage jobs at startups.

Is number of objects true indicator of Amazon S3 growth?

2008-03-23T18:05:00.000-07:00

In my last blog post, I estimated the data stored on Amazon S3 in exabyte range using 18 billion objects stored reported by Amazon CTO, Werner Vogels in his blog post.

In retrospect, it was an over-estimation by several order of magnitude (my bad) that was promptly corrected by MikeDoug using another data point AWS revenue. MikeDoug estimated (comment excerpts below) the data stored to be in 20PB (petabyte) range, way short of my estimates and may be more closer to reality.

No, doubt, it is still a significantly large number for a service that is only few years old. But, S3 growing up fast may not be as obvious from growth in stored objects as Vogels would like us to believe.

A recent report puts ALL of AWS at the 50 to 70 million in revenue for the year.

Let us pretend that, of the 70 million, 40 million in revenue was attributed to S3 alone for last year. That would be $3,333,333 a month for S3. This converts to 22,222,222 gigabytes, or 0.02 exabytes.

Other interesting tidbits if S3 has 20PB of stored data, 18 billion objects and 330,000 registered developers:

On average, each object is only storing about a megabyte of data. This number seems quite low so either deleted objects are being included in the published number of objects or developers are keeping object size low to prevent transfer timeouts.

On average, each developer is only storing 54GB of data. Considering some services like SmugMug are storing terabytes of data on S3, most probably there are lot of registered developers either not using S3 actively for storing data or have services under development.

How much data is in Amazon S3?

2008-03-19T21:49:00.000-07:00

Today, Werner Vogels mentioned in his blog post Happy Birthday, Amazon S3! about the second birthday of Amazon S3 and also shared that by Jan 2008, S3 is storing 14 billion objects. I am not sure why Werner and others at Amazon are so cagey about sharing actual storage capacity used in AWS. In the past, I also have met with either silence or "trade secret" or "competitive advantage" response to my inquiries.

In my opinion, it only creates room for speculation as I am going to do with this post. So, how much data is stored on S3?

My initial guesstimate for stored data volume is between 14 and 70EB (Yes, EB is Exabyte) based on the published information about the size of individual object being one to five GB. Doesn't it seem very high? At first, it did to me. I have been trying to come up with alternate methods to estimate stored data volume like the typical size and type of data being stored by various services that are using S3. Even with an average value of 100MB per object, the stored data volume comes out to be 1.4 Exabyte, still a huge number for such a young service.

What is your estimate? Any suggestions on estimation method to arrive at more accurate number for data volume stored on S3.

Considering that S3 may be hosting Exabyte or more of data with in two years of existence, no wonder all established vendors EMC, IBM, HP and Dell are salivating on getting a piece of the "Cloud Storage" pie.

Bandwidth, one hurdle in adopting Cloud Storage

2008-03-16T23:39:00.000-07:00

This weekend, I read NY Times article Video Road Hogs Stir Fear of Internet Traffic Jam.

Last year, by one estimate, the video site YouTube, owned by Google, consumed as much bandwidth as the entire Internet did in 2000. …

In a widely cited report published last November, a research firm projected that user demand for the Internet could outpace network capacity by 2011. …

Moving images, far more than words or sounds, are hefty rivers of digital bits as they traverse the Internet’s pipes and gateways, requiring, in industry parlance, more bandwidth.

While reading the article, it occurred to me that isn't bandwidth going to be the main hurdle in adoption of storage in the cloud. When clients are not happy with 10/100/1000Mbps connection with application/server/data center, how can they be happy with DSL/Cable/T1/T3 connection to the cloud? I am sure everyone has felt the pain of trying to transfer large datasets over the Internet.

If you review the introduction and growth of various Amazon Web Services (AWS), a comparatively established cloud player, you will notice very limited use cases of Simple Storage Service (S3) on its own with clients outside the cloud. Most S3 usage is fronted by another AWS in the cloud such as Elastic Compute Cloud (EC2). Such combinations overcome the challenge of transferring large amount of data between storage cloud and an application/server outside the cloud over Internet. For cloud storage to be successful, it need to be in the same cloud with application/server or connected to application/server cloud with high speed link.

Any technology that can reduce the data transfer between the cloud services and clients outside the cloud will be the big beneficiary in this trend. Caching, Compression, and Data De-duplication will most likely benefit in the near term. And, the future seems to be very much like the past aka mainframe - Desktop Virtualization, Streaming, and On-the-Fly Visualization.

So, how will new cloud players like Nirvanix, EMC Mozy and Rackspace differentiate?

Are you using online storage services and how?

2008-02-10T18:41:00.000-08:00

Last week, Ethan Oberman alerted me to his online storage service SpiderOak after coming across my post Online Backup Services - What's Next?. Since my post last year, I was contacted by several online backup and storage service providers.

Ethan highlighted differentiation of his service primarily in the area of file versioning, delta transfer, secure sharing across machines and users, and zero knowledge security.

Our approach to online backup and storage varies greatly from our competitors - creating a personalized network concept as opposed to simply online backup. …

Similarly, last year, Marcus Hartwell introduced me to Diino service that also focuses in the area of online backup, storage and sharing.

Most online storage services, since late 90's, are mainly focused on serving one or more activities in data management:

Backup,

Sharing, and

Access.

These services are primarily targeting consumers and small businesses, a bottom up approach with hopes that over time mass adoption will result in acceptance by enterprise IT departments. Strangely, none have been able to make significant impact and gain wide-spread momentum. As previously mentioned, dozens of them have come and go, and I am sure you noticed this trend too.

Though, I did try out several services for a short time, I just couldn't see any becoming part of my daily online routine. And, the main adoption challenges seems to be that either I need something that operates "invisibly" or integrates with my current tools and online activities.

Are you using online storage services and how?

Isilon rebounding this year?

2008-01-28T23:33:00.001-08:00

Earlier this month, a Seattle PI blogger John Cook asked local VCs about the company that will have a breakout year in 2008. Answers by two local VCs, Bill Bryant of DFJ and Jon Staenberg of RCP caught my attention as both suggested Isilon Systems to have breakout year. I don't know their reasons but I hope these VCs are right. As one of the few local storage companies in Seattle, I would like to see Isilon succeed too.

BTW, how do you define and measure a "breakout" year for a company?

There are no doubts about desirability of Isilon clustering technology and the growth of the targeted media and entertainment segment. But in the short-term, Isilon need to cleanup the mess to benefit from this market opportunity, sooner the better. There are three main hurdles with rebound of Isilon - Financial concerns, impact of loss of revenue from key customers and product quality/service issues.

Financial Concern

The financial concerns about Isilon are primarily resulting from delay in 10-Q filing, bane of being a public company. Also, Isilon didn't do a great job of managing the market expectations, for example releasing the bad news slowly (financial restatements, executives change, revenue shortfalls, delay in 10-Q filing). Typically, public companies release all bad news in one shot, take a big hit in the market and then move on instead of slow bleed in the market.

I wonder, like DGM, how many future customers are sitting on the fence concerned about long-term financial viability of Isilon.

And then there's the other problem - googling for technical information I came across a whole set of entries suggesting that there might be some financial problems in the parent.

Loss of Key Customer

As mentioned in 8-K filed Nov 4, 2007, it seems Isilon has some issues accounting for sales to certain resellers and customers. Also, as mentioned by Seeking Alpha, one of Isilon's largest customer Kodak accounted for no revenue in Q3 of 2007.

Company blamed one of its largest customers - Kodak (EK) (17% of revenues last quarter went to 0% this quarter - ZILCH) amongst weakness in Europe for the short-fall.

How much of the impact does these two events have on the bottom line going forward? I decided to take a look at Isilon revenue with and without two of its largest customers - Eastman Kodak and Comcast for last few years.

A quick note about the assumptions and trends in the above charts:

No revenue from Kodak in Q3'07

No change in % quarterly revenue contribution by Comcast in Q3'07 from previous quarter and same % as in Q3'06.

For the quarters, where % revenue contribution by Comcast or Kodak were not available, estimates are made based on % annual revenue contribution.

Closer the - Kodak or - Comcast curves to Total curve, lower the contribution by those customers to total revenue.

Since IPO, Kodak had contributed a large slice of total revenue to Isilon. But, the past history also indicates that in second half of year Kodak typically contributed less to total revenue compare to first half. In my opinion, Kodak and Comcast enabled Isilon to go public at least two quarters earlier than they should have. Is the disclosure of no revenue from Kodak in Q3'07 indicate that future revenues from Kodak are likely to be significantly low? Lets hope that independent review of some sales is not going to impact revenues from Comcast.

Service/Product Quality

As any product company selling to large enterprise learns sooner or later that an internal post sales service organization complementing third party service providers is needed to provide exceptional service. It seems recently Isilon has been building up its internal service capabilities.

The change in contract manufacturer seems reasonable considering agreement was close to expiration though I expect there will be short-term hardware quality pain during the transition and ramp up. I couldn't find any smoking gun to support any talk about product operation or compatibility issues.

Summary

My take is that addressing financial concern should be Isilon's top priority. Clearing the financial picture will be the main hurdle in their rebound. It is also great to see Isilon returning to its roots in technology and product innovation instead of trying process innovation like established companies.

Data Storage Companies in Seattle

2008-01-21T17:46:00.000-08:00

Recently, during review of Seattle tech ecosystem (most don’t target infrastructure segment), it was interesting to note the diverse, emerging and unique, areas in data storage being targeted by local tech companies - Isilon Systems, F5 Networks, Bocada, Amazon and Microsoft.

Whom did I miss on this list of local data storage companies? Can Illumita, a local “virtualization over Internet” startup, be considered as a potential influencer in data storage space?

This year, I plan to write more often about these and other local Seattle data storage companies.

EMC Academic Alliance - A Good Storage Education Initiative

2007-12-25T15:08:00.000-08:00

… Continuation from last post, Are We Educating Future Storage Professionals?

I subscribe to ACM Digital Library and enjoy reading the latest storage research happening in academia and research organizations. Recently, I came across an interesting paper in Proceeding of the 8th ACM SIG-Information Conference on Information Technology Education on EMC Academic Alliance program (See, Storage Technologies: An Education Opportunity, Ed Van Sickle et. al. SIGITE’07, October 18-20, 2007, Destin Florida USA).

In this paper, Ed discusses EMC realizing during hiring process that very few recent graduates had any knowledge of storage technologies. Initially EMC tried boot camp approach. Then, EMC concluded that greater benefits may be achieved by creating courses focused on storage technologies at university level.

This gave birth to EMC Academic Alliance Program [PDF] with goals of educating CS/IT students on storage and support for knowledge transfer, guest lectures, and site visits. In my opinion, it is an impressive initiative and kudos to EMC for recognizing the shortage of storage skills and taking the lead with potential solutions. Why are EMC bloggers not highlighting and promoting such a positive initiative?

Why is an industry association like SNIA not leading Academic alliance initiatives? It doesn’t look like education is a part of their new mission.

The paper also showcases the implementation of this initiative at four universities and provides overview of the courses held and plans for future classes. This is a great compliment for the course offering under this program at Penn State University (PSU).

Subsequent offerings of the course filled to room capacity based on the positive word-of-mouth that the course generated. In fact, students from other PSU colleges (engineering, computer science) have requested to be added to the course.

Of course, program also has its challenges as encountered at University of Massachusetts, Dartmouth with course material being unsuitable for the targeted student segment and unavailability of suitable text book, and at North Carolina A&T University with low student interest and course enrollment.

The paper also mentions the interest of Dr. Cameron (one of the co-authors) at Penn State University in developing a three course storage track but being constrained by lack of teaching resources. Hopefully, he can attract other storage vendors to fulfill the vision of a storage track and overcome the lack of teaching resources through guest lectures by industry professionals.

I wish paper had further explored student/instructor survey results and the challenges facing the program.

Will EMC collaborate with its customers, partners and other storage vendors in growing this initiative? How can rest of the storage industry help in growing the program? How can storage bloggers help?

Are We Educating Future Storage Professionals?

2007-12-23T11:08:00.001-08:00

Happy Holidays to all readers.

During my visit to India over the Thanksgiving, I met some friends and family members whom I haven’t seen for a long time. They inquired what I do and when I responded that I work in data storage, the invariable follow up was “What kind of database?” ;-)

I am always interested in learning more about various storage technologies irrespective of their relevancy to my job at that time. So, when I moved to Seattle, I decided to check out the course offerings in storage by local universities and colleges. To my surprise, there was not even a single class offered on data storage at any of the local educational institutions including University of Washington.

What the above two examples have in common is the lack of awareness in data storage. Despite the criticality of IT infrastructure and data storage to corporations, there is lack of knowledge and focus on these topics by most IT professionals, whether experienced or recent graduates. Most of the storage knowledge seems be gained through on the job or vendor training that typically focuses on only working with specific products. Educational institutions also seem to be oblivious to the need of educating storage technologies to their CS/IT students.

Most storage vendors offer training on their own product portfolio with little or no focus on underlying storage technologies that make up their products. SNIA has tried to bridge the gap through their education tutorials at SNW and vendor-neutral storage certifications. If my experiences at spring SNW is any indication, most attendees to these tutorials are storage professionals themselves. The certifications are also targeted at validating the skills of IT professionals already working in storage rather than attracting experienced IT professionals from non-storage domains.

Is this apathy by storage community toward storage education resulting in storage skills gap? Is storage community doing anything to bridge the skills gap in storage knowledge of experienced IT professionals as well as recent CS/IT graduates?

In my follow-up post, I will discuss an interesting initiative taking place to address the gap in storage education at academic institutions.

Is your encrypted data also RIP with NeoScale?

2007-12-12T22:15:00.000-08:00

There has been lot of talk in trade publications and blogs about demise of NeoScale. Jon Toigo asked interesting questions about the go-forward strategy of nCipher after picking up remains of NeoScale. Storagezilla also air-raided the encryption appliances by poking holes in Decru and his reluctant acceptance of appliance approach to some "data at rest" encryption problems (I wonder what problems, he is referring to).

After hearing the demise of NeoScale, my second reaction was:

"Hmm ... I wonder if NeoScale customers will think about decrypting the terabytes of vaulted data that was encrypted using Cryptostor before their appliance fails and no chance of finding a replacement."

I am sure Decru and other encryption vendors are salivating on the opportunity to sell in to NeoScale customers, BUT

Can their encryption solution decrypt the data encrypted through Cryptostor?

That is the question NeoScale customers should be asking when talking to encryption vendors about replacing Cryptostor.

I expressed my concerns to some people who are using encryption products. None had considered and/or planned for decrypting the data upon losing access to the tool (product) or method (algorithm) that was used to encrypt the data. It is a real scenario for encrypted data on any kind of removable media despite availability of correct encryption keys.

Just imagine what will you do if seven years from now a government agency requests financial data that was encrypted and archived on a removable media vaulted offsite. And, you realize that you can't read data because you no longer have the original system capable of reading and/or decrypting that data. I experienced the same challenge in a customer environment few years ago though with unencrypted data.

Unlike Mark, I am not very enthusiastic about encrypting "data at rest" specifically where encrypted data is stored separately from the system that wrote the data or is capable of reading that data. The demise of NeoScale may be just the wake up call for the trouble you may get into if you encrypt the "data at rest" and you have no way to decrypt the data because you lost the method or tool or keys to decrypt.

Is NetApp the only player in File Systems & Storage?

2007-12-01T13:16:00.001-08:00

If you read September/October 2007 issue of ACM Queue, you may get that impression.

Recently, I read this issue and interestingly all three articles in Q focus on File Systems and Storage covering pNFS, Hard Disk Drives and Storage Virtualization topics are authored by people from Network Appliance.

Though, these articles are worth the read, it most probably is the oversight of ACM Queue editors that focus on File Systems and Storage became focus for Network Appliance.

Only saving grace for the editors is the Interview with SUN engineers who created ZFS else NetApp marketing will be distributing reprints of whole issue.

Storage Challenge for Media Firms: High Definition

2007-11-29T22:55:00.000-08:00

My last blog post on help designing a suitable storage solution for a creative media firm is generating some quality comments from lonerock, Paul Clifford, and others. Unfortunately, due to few other commitments, I wasn't able to further communicate with the executive at the creative media firm. Below is our last communication that highlights further details of his environment and the challenges faced by his firm.

Why are you considering moving from NAS boxes to SAN solution? Are you using NAS boxes from a specific vendor?

JD: Whenever we have needed to increase our storage we have usually ended up buying more NAS heads – which means we need to manage more equipment. Our equipment is purchased from Dell and is our preferred hardware vendor. Earlier this year we had anticipated growth and increased our storage at that time by 200%. We purchased 2x 2950 Dell Storage Servers with Microsoft SS 2003 giving us about 2.5TB of space on each. We used one of the servers for a live backup. Each of these machines are dual-quad core intel. Middle of the year – we are running low! So instead of buying a NAS, we are looking at the Celerra (though we will use it as a NAS head) as it seems to have a better scalability path. We would then use our current 2950 as render servers.

Are you able to share more details on what specific storage hiccups and access speed issues you are facing?

JD: We do a lot more HD animation work than before which is leading to larger render files, also some our client projects now tend to go on for a year or longer. And we need to keep them active on our systems. Our reluctance to remove old render files, tend to increase size dramatically over the lifespan of a project. Over long weekends, when our render farm is churning away files, we need to make sure we have a clear 200-250GB of space – but usually we are struggling and end up spending a lot of time pruning projects to clear space. Speed wise we don’t think we are in a bad shape – though we have 40 users hitting our NAS box, and it the monitoring device does show a 100% utilization and queing of requests. We were thinking of breaking in more NAS device and breaking up user groups based on that – however again we end up with more equipment and management issues.

How are workstations, rendering servers and NAS boxes currently connected? Can you provide further details of your current infrastructure?

JD: All connections are via a gigabit Ethernet, we have Cisco catalyst switches. Our current storage server is connected via a dual fiber link to the switch. Users typically open 3D files which contain 100 or more linked texture and material files, they work on these files and will typically send them over to our renderfarm using a render manager. The render servers pick up the request and start the rendering process, depending on the scene it could take anywhere between 20 – 40 minutes for the frame to be created and written to disk. Each frame can be 2-3mb in size. We usually do multiple passes, that is 5+ frames make up 1 frame. 30 frames = 1 second of animation, so it adds up!

What are you finding attractive about NS20 Celerra (another NAS head)?

JD: Scalability mainly and the ability to easily manage storage and dynamically carve space depending on requirements, we will initially go with 8TB of usable space and add more drives trays as and when we need them in future. Also, the backup system for celerra with their snapshop option looks pretty impressive.

What is a typical day to day workflow that utilizes and strains the current infrastructure?

JD: As our 3D visualization and rendering is our core, our biggest issue has always been render capacity and storage capacity. We are in good shape with our render capacity about 20+ servers, our storage is the current issue, we can render more in a shorter time frame, also our HD frames are twice or thrice the size they used to be. So there is a constant pruning (which means we may delete files we really shouldn’t be!) and a trying to archive projects to tape the moment they are done only to bring them back on again when a client comes back in a few months with changes.

Design a Storage Solution for a Creative Media Firm

2007-10-09T21:04:00.001-07:00

One thing, I enjoyed the most in last four years while writing this storage blog, is the interaction with professionals from diverse background discussing wide variety of issues. Time to time, I was approached by readers looking for new opportunities or to fill open positions in their organizations. No, I am not setting the stage for launching a classified service on my blog.

The reader communication that attract me most are the ones where I learn something new or I can extend help to others or introduce two people together. And, such contacts didn't stop despite my absence for past few months.

A product development executive at a creative media firm contacted me a month ago requesting advise on storage solution for his firm. Unfortunately, I wasn't able to communicate regularly after initial contact. In our last conversation, he agreed for me to post his initial email (redact identifying information) on my blog.

Can you advise on a suitable storage solution for him? What questions will you ask to get additional information? What suggestions will you make?

Dear Anil,

I was reading through storage solutions and came across your very informative blog. I am not sure if you can help me on this; we are looking to deploy a SAN solution in our company and since my knowledge on this topic is only building up now I am a bit shaky about going down this path.

Just a quick overview – we are a creative media company with about 40 users and growing fast, our work involves 3D visualization, video production and interactive media development targeting architecture and pharma companies. We are continually fighting a battle with our storage requirements, we have generally been adding a NAS box as we needed more and using the older NAS boxes for admin usage etc. Our main hiccup is storage and to a lesser degree speed. We don’t have any serious application servers, just workstations and render servers in our production pipeline. We are looking at the NS20 Celerra from EMC, but am a bit unsure of this path and it’s applicability in our environment.

Thanks

JD

Thanks for sticking around!

2007-10-01T22:42:00.000-07:00

It has been quite a while since I last blogged. This summer had high-and-low stored for me at personal level. The summer started out great with family get together and celebration but ended with death in the family.

Blogging and Storage weren't something I though about much in last two months. As someone said, life goes on, it is time for me to move on from personal tragedy. Thank you everyone for inquiring about my absence from blogging and whereabouts. I hope to slowly ease back in to blogging again.

If you are reading this post, thanks for sticking around!

Just Say "Thanks for Feedback"

2007-08-17T22:45:00.000-07:00

For almost a month, blogging and blog reading took a back seat to hosting friends and family from four different countries, playing golf and Wii (addictive!), rigors of the day job and enjoying the summertime.

I most probably will still be on blogging hiatus for few more weeks if not for the battle of the words among storage bloggers Robin Harris, Barry Burke, and others. I guess I got embroiled by making an observational comment on Robin's Blog post resulting in pen-lashing from Storagezilla.

Overall, I am very surprised with the emotional responses and the offense both bloggers and commentators are taking. Be nice, boys! There is some great feedback in these blog posts and comments for everyone to take home. Whether, like it or not, just saying thanks for the feedback, will go a long way.

After storage blogging for almost four years, I would be saying thanks for the feedback, if I was receiving such thoughts about my blog from someone. May be, I am not cut out for EMC DNA mutation. BTW, thanks Chuck for acknowledging the storage trenchtrolls.

This war of words also reminded me of an incident, I was involved in, about two years ago. At that time, Hu was a class act in responding to my criticism of his blogging. In retrospect, I was out-of-line. Despite this incident, I was invited and hosted at HDS Executive Briefing Center by Jeremiah, previously HDS community evangelist. For this very reason, I admire and have great respect for both Hu and Jeremiah.

My offer to storage bloggers, readers, industry insiders and outsiders remains same. If you ever visit Seattle area, get in touch. I will love to pick your brain over a beer/coffee/lunch/dinner. I will do the same when I am in your town, just send me your contact info.

Of course, any feedback through comments, email and phone is always welcome too.

Tales of Storage IPOs: What happened to Quality?

2007-07-24T00:32:00.000-07:00

A comparison of upcoming IPOs, Netezza, BladeLogic, Voltaire and Compellent, with recent storage IPOs. If VMware is to be shown on this chart, it will be outside of upper-right quadrant.

Power Consumption by Google Services

2007-07-16T21:28:00.000-07:00

Even though, Google doesn’t share a lot of details of their infrastructure, as we have seen from limited published information, they are obsessed with continuously monitoring, managing and improving the efficiency of their infrastructure.

Recently, Robin Harris attended the Google conference on scalability and then mused How Yahoo can beat Google. Few months ago, Google published results of their work on disk drive failure in paper Failure Trends in a Large Disk Drive Population [PDF]. It was extensively covered in blogosphere including by me in blog entries SMART not so smart in predicting disk drive failure and Google Findings of Disk Failures Rates and Implications and by Robin Harris in his blog entry Google’s Disk Failure Experience.

Google has done it again and presented results of their work on power consumption and provisioning in paper Power Provisioning for a Warehouse-sized Computer [PDF] at the ACM International Symposium on Computer Architecture, San Diego CA, June 9 – 13, 2007. In this work, Google researchers, Xiaobo Fan, Wolf-Dietrich Weber and Luiz Andre Barroso looked in to 15,000 servers running three different applications – Websearch, Webmail and Mapreduce for six months to determine the power usage characteristics at Rack, PDU and Datacenter levels.

Google Services

Websearch: A service with high request throughput and large data processing for each request.

Webmail: A disk I/O intensive service. Machines configured with large number of disk drives. Each request involves a relatively small number of servers.

Mapreduce: A cluster dedicated to running large offline batch jobs. Involve process terabytes of data using thousands of machines.

Key Findings

The key findings from this work are:

The difference between maximum power used by large number of computing devices, cumulatively, and their theoretical peak usage can be as much as 40% in datacenters.

It may be more efficient to leverage power management techniques at datacenter level than at rack level.

Nameplate ratings are of little use in power provisioning as they significantly overestimate actual maximum usage.

CPU utilization as a measure of machine-level activity produces accurate results for dynamic power usage especially with large group of machines. The dynamic power range is less than 30% for disks and negligible for motherboards.

Using maximum power draw of individual machines to provision the datacenter, will have some stranded capacity.

A mix of diverse workload reduces the difference between average and peak power, an argument in favor of mixed deployment.

Idle power is significantly lower than the actual peak power, but generally never below 50%.

CPU dynamic voltage/frequency scaling may yield moderate energy savings (up to 23%) at datacenter levels.

Peak power consumption at the data center level could be reduced by 30% and energy usage could be halved if systems were designed so that lower activity levels meant correspondingly lower power usage profiles.

I believe there is a value proposition in this study waiting to be discovered by Copan Systems for their MAID technology to be applied at PDU and datacenter scales or even along the lines of SUN Project Blackbox.

More details from this study later.

Tales of Storage IPOs: Hardware or Software and Components

2007-07-15T01:30:00.000-07:00

Recently, an interesting question about the preferred packaging of the storage product by startups came up during my conversation with few private equity acquaintances. The opinions were kind of biased and fell along their past professional background.

The opinion from finance-focused guy was that most software products require less initial investment in infrastructure and working capital needs for distribution, service and support. The software startups tend to become cash flow positive and profitable lot more quickly with early market successes. The opinion from sales-focused guy was to have a product in physical form that a potential customer can visualize, feel and touch. A physical product tend to raise fewer investigative queries about the inner-workings and the inquisitive knob-turnings that tend to happen with software product.

I am of the opinion that irrespective of the storage product packaging, the differentiation tend to come, most times, from the software running under the covers or as a standalone product. One lesson, I learned from my last entrepreneurial storage adventure was that during early round of financing, the pressure for a "working" prototype is lot greater with a storage software product than hardware product.

I decided to further look in to the question of whether startups with physical storage product are perceived better by looking at the value placed on them in the marketplace. Unfortunately, the earliest available public financial data for startup is when it files for IPO registration.

This is an attempt in finding patterns for startups from the public financial data of recent IPOs by Isilon, Data Domain, Riverbed, Mallanox, Commvault and Double-Take. A visual representation of revenue, income/loss, market capitalization and change in market cap after 12 days of trading for these six recent IPOs is shown below.

Be mindful of the limitations and assumptions made to simplify the analysis such as few data points, number of outstanding shares remaining unchanged at IPO, market cap as indicator of value of company and equating price change with change in market cap over short period of trading after IPO, and adjusting revenue and income to make them consistent across all IPOs.

Some of my observations from this analysis are:

Storage software and component vendors were valued less at IPO compared to storage hardware despite being profitable and revenues at par or higher. During early trading, public markets didn't reward software and component vendors for better financial strength either. The gains in market cap during first 12 days of trading were significantly higher for hardware IPOs. Being profitable doesn't necessarily gets rewarded. The extent of profit/loss had little bearing on performance of storage IPOs. Maturity of business results in lower rewards. Double-Take was seriously undervalued and under-appreciated in all aspects.

Last Friday, Gary also wrote about the price performance of storage IPOs during past six months in his post IPO Class of 2006. He mentioned that Riverbed and Double-Take had highest percentage price gain, while Isilon is swooning and Commvault is in doldrums. May be the efficiency of the marketplace is being finally realized.

As I was using IPO data as a proxy to startup value, I decided not to extend the time period of analysis (may do anyway, now). The primary reason was to simplify the analysis and keep the data mining and normalization workload manageable - several vendors had secondary offerings, number of outstanding shares may have changed, temporary changes in financial data resulting from IPO event and proceeds, and lock-up expiration.

Share your thoughts and opinions on the success/failure of startups and storage IPOs.

Discontinuing 3PAR posts

2007-07-14T16:46:00.000-07:00

I have decided to discontinue further discussion of 3PAR due to underwhelming response by 3PAR to my request for technical information beyond marketing brochure. Check out the blog post 3Par Data: Utility Storage by Ben Rockwood for his opinion on 3PAR, though I disagree with some of the points raised in his post.

Also, check out the following 3PAR coverage in blogosphere:

3PAR, data storage virtualization company readying for IPO by Matt Marshall

David Scott, 3Par CEO on Thin Provisioning Gotchas by Marc Farley

3PAR: Innovative or Par for the Course? by Jon Toigo

3PAR: System Design. Part I

2007-07-09T22:23:00.000-07:00

One of the core features of 3PAR, in my opinion, is System Design that facilitates scalability, availability, performance and ease of use. As posted in my previous blog entry, Craig Nunes mentioned two key attributes of their system design, scalability and clustered modular architecture.

… our customers have taken great advantage of the scalability of the array. The clustered, modular architecture eliminates the price premiums of the monolithic arrays and scaling complexity of modular array architectures.

Another reader mentioned two key benefits of 3PAR system design, the performance and ease of use, similar to what I heard from several users and evaluators of 3PAR at SNW. The data is spread across every spindle, up to 2,560 drives, and workload is spread across all active/active controller nodes. The LUNs are built from available raw disks with little need for pre-planning, pre-configuration of disk/parity groups or deciding on what disk spindles to use.
3PAR Utility Storage product brief also gives some insight in to the system. I know the brief is a marketing spin but I am working on getting more technical details from 3PAR.

Central to the design is a high bandwidth, low-latency backplane that unifies …, … modular and upgradeable components into a highly available, … automatically load-balanced cluster.

A full-mesh, … passive backplane provides a dedicated … data path between each and every 3PAR ASIC, one of which resides in every 3PAR Controller Node.
Physical disks … are divided into uniform 256-MB chunklets. Chunklets from across the system are then automatically selected and grouped to meet user-defined levels of performance, cost and availability (varying such parameters as RAID type, drive type, radical placement and stripe width).

Customers can non-disruptively alter a volume’s underlying RAID protection level, drive type, stripe width and/or radial placement.

In my next post, I will dig deeper in to 3PAR system design, hopefully with some help from 3PAR and their customers so that I don’t need to make up stuff.

3PAR: The Follow-up

2007-07-01T16:36:00.000-07:00

Sometime, a blog post touches the nerves and that was the case with my last blog post 3PAR: Reversal of Fortune. I heard from numerous readers including people from Bluearc and 3PAR. Craig Nunes, VP Marketing at 3PAR sent me a detailed email highlighting the reasons why customers are choosing 3PAR.

The success of our products, as you mentioned hearing from several customers, has a lot to do with ease of use. Those who have chosen 3PAR have been able to provision capacity in seconds without preplanning. They have been able to eliminate performance tuning activities because all volumes have access to all resources within this massively parallel system. They have also been able to move data from one storage tier to another or redistribute existing volumes across new resources as they are added to the system in a single command, completely non-disruptively.

That being said, ease of use is only part of it -- efficiency is a also a big part of why customers have deployed 3PAR in their data centers. As you mentioned, Thin Provisioning is getting a great deal of attention and is now moving into large-scale deployments within very conservative companies. But efficient local and remote copy capabilities -- both built on thin copy technology -- are another major reason customers have chosen 3PAR. Organizations have taken substantial advantage of these products, deploying hourly (writeable) snapshots for granular and rapid recovery, and cost-effective tier one remote data replication solutions without the need for vendor professional services.

Finally, our customers have taken great advantage of the scalability of the array. The clustered, modular architecture eliminates the price premiums of the monolithic arrays and scaling complexity of modular array architectures. Our customers start small, with as few as two controllers, yet grow non-disruptively and massively within a single, fully-tiered system. Organizations, especially those with high growth requirements, have chosen to grown within their 3PAR array instead of facing the complexity of managing 5 or 10 or more arrays from other vendors.

Another reader put things more succinctly.

Thin provisioning is great, and is definitely a killer feature, but it's not THE killer feature. I'd say it's one of four things that 3PAR does really well. The combination of those four things in one box, is what differentiates 3PAR from other vendors in the market.

The key claims, Craig made were:

Ease of use - Provision capacity without preplanning. Eliminate performance tuning.

Efficiency - Thin provisioning. Efficient local and remote copy capabilities.

Scalability of array - Clustered modular architecture. Grow non-disruptively and massively within a single fully-tiered system.

I don't have any first-hand experience with 3PAR product. Do you have any experience with them? What is your opinion about 3PAR product and technology? This month, I would like to further look in to 3PAR technology. Any documents, comments, emails and phone calls that help me understand 3PAR better are most welcome.

3PAR: Reversal of Fortune

2007-06-25T21:18:00.000-07:00

Last weekend, while browsing my archive on the mobile drive, I came across a text file with interesting quote. I don't know who said this and where it was published. But the timestamp on text file shows August 2002.

3PAR won't make it as far as BlueArc did, but will mostly likely fail for many of the same reasons. A box is a box is a box...."What's the price per megabyte?"....This is unfortunate, but what customers really want and need is faster, better, cheaper storage WITH integration with all the other pieces of the SAN and applications. Building a bigger, faster box takes a while. Certifying and integrating useful applications, host support, switch support and the ENDLESS list of combinations with this and HBAs takes a LOT longer. Not to mention costly and engineering-intensive real-world performance testing to prove it really is better.

As with many things, if it's not faster, cheaper and better, there isn't much motivation for large customers to take the risk; especially in this climate.

~~Write a comment or send me an email if you know the origin of above quote.~~ Update: Anonymous comment pointed to B&S Message Board as the source of the above quote.

Five years later, I don't know about BlueArc but 3PAR seems to be on its way to becoming a successful established subsystem vendor.

At SNW, I heard praise for 3PAR by several customers who were using 3PAR subsystems and also prospects who were in evaluation phase. What was surprising that not one of them mentioned much-hyped thin provisioning to be the primary reason for selecting 3PAR. All pointed out the 3PAR volume manager and striping of data across available disk resources being the primary reason with comments like "HP EVA like capabilities in 3PAR go far beyond EVA."

Both, 3PAR marketing at SNW and CEO David Scott in Byte & Switch, seem to be highlighting thin provisioning as the main reason for their success in highly competitive subsystem market with very conservative large enterprise customers. Is it really so?

What are your reasons for selecting or not considering 3PAR subsystem?

Gear6 trailblazing Network Caching

2007-06-20T00:16:00.000-07:00

Earlier this week, I had great conversation with Gary Orenstien and Jack O’Brien at Gear6. Here are the excerpts from our conversation.

How is Gear6 doing?

Gear6 seems to be doing well. Several units are currently in field being evaluated by various customers. No specific number of units provided, just a wide range between 10 and 100. Company has over thirty employees and financially all set in the near term. Company has started to focus CACHEfx on financial analytics, energy and animation segments and will expand focus by the end of the year.

What are the benefits of network based caching?

Network caching enables increased cache utilization, flexibility and scalability. Caching is moving from end devices to network and becoming a network resource.

What one factor is attracting customers to your caching solution?

By nature of caching, the obvious benefit to customer is performance. Most customers who come to Gear6 have performance problems, variable workload and demand certain Quality of Service. The success rate is very good with evaluations by customers as CACHEfx appliance doesn’t require forklift replacement.

How is Gear6 doing caching?

CACHEfx appliance doesn’t use any conventional mechanical disk storage internally, 100% RAM cache and is pass through to persistent storage. Robust single purpose appliance designed to do one job and do that job very well.

The caching is performed intelligently. The intelligence focus on how and where data is placed within the appliance. There are extensive built-in statistics. Most customers are impressed by network sniffer like capability.

In the past, cache was a constrained resource. Now, focus is on right-sizing cache. CACHEfx expands from quarter TB to multi-TB, can be preloaded with data from persistent storage and adjust to variable I/O profile.

What are the reliability, availability and scalability features of CACHEfx appliance?

It is a clustered appliance, scalable from quarter TB to multi-TB. The appliance can be expanded on the fly. Also, appliance only acknowledges writes only when persistent storage sends acknowledgment.

Is the CACHEfx installed at D.E. Shaw working with Solaris cluster?

Gary declined to comment on infrastructure details of customer. He claimed customer pleased with the solution.

Any plans to introduce network caching for block-level traffic? The present product seems to focus on NFS only.

The present focus is on NFS, market is large enough. The sweet spot is where customer is using 100+ concurrent clients accessing single dataset, most tend to be NFS. No firm plans for addressing CIFS or block level traffic. The primary industry focus on financial analytic, energy and exploration, electronic design, animation, biotechnology, and media, primarily HPC oriented tasks.

How does network caching stack up with parallel file systems and clustered storage?

Caching addresses I/O constrained systems rather than processing constrained. Parallel file systems and clustered storage solutions are capacity centric not performance centric, providing global namespace for ever expanding storage capacity. They are not low latency solution. Network caching is a complementary solution, capacity complemented by performance. Gear6 solution complements Netapp OnTAP GX, IBRIX, Isilon and Acopia.

Do you have any thoughts on potential application of CACHEfx in a Wide Area Filer Network environment?

The CACHEfx has enormous potential in variety of environment. But we are currently very focused on solving customer problems within the data center. We are open to partnerships in other areas.

Bountiful Bandwidth Lagging Latency

2007-06-17T13:15:00.000-07:00

Recently, I came across an interesting article published in 2004 comparing growth, reasons and handling imbalance between bandwidth and latency. Excerpts below are from Latency Lags Bandwidth, Recognizing the chronic imbalance between bandwidth and latency, and how to cope with it. By David A. Patterson, Communications of the ACM, October 2004/Vol. 47, No. 10.

In the time that bandwidth doubles, latency improves by no more than a factor of 1.2 to 1.4.

Reasons for Bountiful Bandwidth

“There is an old network saying: Bandwidth problems can be cured with money. Latency problems are harder because the speed of light is fixed – you can’t bribe God” – Anonymous.

Moore’s Law helps bandwidth more than latency.
Distance limits latency.
Bandwidth is generally easier to sell.
Latency helps bandwidth.
Bandwidth hurts latency.
Operating system overhead hurts latency.

Coping with Lagging Latency

Caching: Leveraging capacity to help latency.
Replication: Leveraging capacity to again help latency.
Prediction: Leveraging bandwidth to again help latency.

Marketing Latency Innovations

The difficulty of marketing latency innovations is one of the reasons latency has received less attention thus far.

Perhaps, we can draw inspiration from the more mature automotive industry, which advertises time to accelerate from 0-to-60 miles per hour in addition to peak horsepower and top speed.

Where do you focus, Bandwidth or Latency?

2007-06-12T20:07:00.000-07:00

Since my first post about Gear6, Gary Orenstein and I have been exchanging emails discussing various aspects of storage caching and Gear6. Recently, he commented in response to my request for pointers on storage caching market and implementations:

When I find interesting items related to caching I usually post on our blog. The thing is, there really hasn't been anyone promoting network-based caching until Gear6.

With rising interest in flash memory and SSDs, I am finding storage caching quite intriguing. I decided to start from basics.

What problems does caching solve?

The major benefit of caching is in reducing the latency whether caching is part of the web, network, file system, storage device, processor or memory. What is latency? Any delay in response to a request.

Bandwidth Bias

One consistent theme struck me odd as I started studying caching is how often we suggest more bandwidth as a solution to the slow performance issues and how little focus we give to the latency side of the problem. What is bandwidth? The amount of data carried from one point to another in a given time.

Even in iSCSI world, we all hear how 10GbE will be the inflection point, indirectly giving the impression that bandwidth is the bottleneck in iSCSI adoption. What is the real bottleneck in iSCSI? Is it bandwidth or latency?

I guess it sounds more impressive "With 10GbE, the bandwidth will increase 10X so you will be able to push ten times of data but latency will only be reduced in half (approx)."

From the productivity aspects of users and applications, a predictable and quick response to a request seems to be considerably more important than the amount of data being transferred over a specified period. What good more bandwidth does if data needs to wait for processing? A balance between bandwidth and latency need to be considered in designing solutions.

In the end, my impression is that most of us tend to focus too much on bandwidth and too little on latency.

Wikibon, The Improvements Needed

2007-06-06T23:09:00.000-07:00

As I mentioned previously, Wikibon project is very interesting and schedule permitting, I plan to monitor its progress. I see the value of collective intelligence and bringing down the barriers in market research and industry analysis segment. If the approach succeeds, it will revolutionize this industry, the way Wikipedia did to Encyclopedia business.

The intent of this post is not to dismiss the initiative as another hype of social networking era. All new experiments go through a phase of trial-and-error before finding their footing and niche. I feel Wikibon is currently in that early phase where Dave and his team are trying various things to see what sticks, what not and what will make them realize their vision.

The objective of this post is to help them during this early phase by making two very specific suggestions for improvements.

Lead with Content

Overall, Wikibon started with a good web presence. Only design suggestion will be to lead the presence with content and cleaner interface otherwise it just take away the community and participatory feel of the initiative. Some annoyances:

Too many choices and information crammed in to home page.

Unnecessary and excessive use of text boxes, fonts with different colors and sizes and slide style boxes and graphics.

As Chris Evans commented and I agree that to gain any type of mindshare, Wikibon need to highlight the content not the people.

Do you really need Wiki format?

It is great to see 340 articles on variety of topics already posted on Wikibon. Most articles seem to be "independent" in nature, written by individual authors containing only their opinion with very little scope to modify content by others. I found content to be more fitting for blog format instead of wiki.

This is a typical challenge in most wiki projects. Is the topic and content conducive to modifications by others? If content is more conducive to be commented by readers instead of modification then it is more fitting in a blog format. I am sure you will also be able to see the content that is likely to be modified and added with new information and the content that is likely to receive comments.

Compare the Wikipedia Backup page with following Wikibon pages for Backup articles. It doesn’t take long to identify pages that are more likely to be modified or content added by someone other than the original author/creator.

Implementing fail proof backup and recovery
Backup and recovery options
Backup and recovery techniques
Sizing up backup and recovery options
Data de-duplication and the low-end backup/restore choice

Checkout the storage market prediction trading feature at Wikibon. It is an excellent feature that has potential to leverage the power and knowledge of the community.

Time permitting, I may review Wikibon further.

Wikibon, An experiment in Collective Intelligence

2007-06-05T23:43:00.000-07:00

Few weeks ago, David Vellante contacted me about his new project, Wikibon and invited me to attend Peer Incite research meetings. Wikibon is a project where he is trying to harvest and share the collective intelligence of IT community for market research, industry analysis and insights. Having previously founded the storage research group at IDC, it was no surprise that Dave picked enterprise storage as the first industry segment to target with Wikibon project.

What piqued my interest?

Considering industry analyst world being a walled garden with entry only allowed to chosen few who can pay hefty entrance fee, Wikibon is an interesting experiment. Any cracks in garden walls are a welcoming change for Average Joe like me. But my interest in Wikibon extend beyond just an open source experiment in IT market research. I am more excited about the harvesting and sharing collective intelligence aspects of this 'public' experiment.

How beneficial will it be for an organization to make decisions based on this collective intelligence instead of listening to chosen few with loudest voice or political connections? Unfortunately, considering the competitive advantage such approach offers, few organizations who experimented with collective intelligence internally willing to share and discuss their methods and findings publicly. I believe that blogs and wikis are not just external facing marketing communication tools for enterprises. They also make excellent methods for harvesting the collective intelligence of everyone with in an organization especially those operating in knowledge-intensive industry.

Unfortunately, Dave couldn't make me get up early enough in the morning to attend a meeting at 9:00am ET (6:00am on my coast), later turned out to be just a typo and time zone confusion. Finally, this morning I attended the Peer Incite research meeting on Data De-duplication topic. Even though, this topic doesn't excite me anymore [More in a later post. I have moved on to other exciting and new topics.], the affiliation of vocal participants and dynamics among the participants was interesting to observe.

So, what is my impression and feedback on Wikibon project, community web presence and Peer Incite meeting?

As mentioned before, Wikibon project definitely has piqued my interest irrespective of reasons aligning with Dave's vision or not. I am planning to monitor its progress, share my opinions, and participate and report as time permits.

[Too late in the night] I will try to continue my feedback on this project in another post.

Blogging Hiatus

2007-06-03T17:36:00.001-07:00

Last couple of weeks, I was absent from any blogging due to back-to-back trips to Anchorage and Princeton. Unlike Storagezilla going off the grid, my blogging hiatus was unintentional and due to the demands of the day job and personal life. The highlights of trips were experiencing the scenic beauty of Alaska for the first time, opportunity to play 18 holes at Bunker Hill Golf Course, visiting Princeton University and talking to couple of very smart people.

Note to the readers: Blog posts will be sporadic from June through August due to demands of few other personal initiatives.

SaaS Panel Discussion Recap

2007-05-19T16:38:00.000-07:00

As mentioned before, last Wednesday I moderated a panel discussion on Software-as-a-Service (SaaS) for IIT-PNW at Google Kirkland campus, an amazing experience. Our panel guests represented broad spectrum of SaaS ecosystem and the audience liberally peppered them with questions on wide variety of topics.

Interestingly, only consensus we had between panelists and audience was that SaaS will grow further and will have significant impact on various business and consumer activities.

Defining SaaS. Web 2.0 vs. SaaS. Consumer vs. business focus. SaaS meant different things to different panelists and audience members.

Start small. Target small. Improve quickly and frequently. Generate demand quickly. Scale as you grow. Enable experimentation.

SaaS strengthens and grows further as web access becomes ubiquitous and available on various devices, specifically growth with internet access through mobile devices.

Migration from Software-as-a-Product (SaaP) to SaaS. Benefits of frequent feature enhancements and quick customer feedback. Concerns about accessing data and services offline. Migration from pure web complemented by desktop client option.

Most SaaS growth in application area and little in infrastructure area. But greater and quicker adoption in infrastructure area.

Main benefit of application SaaS in collaborative namespace. Main benefit of infrastructure SaaS in someone else responsible for muck.

Tools and platforms for SaaS development. Doubts about reaching a stage where operating system as SaaS in near future. Concerns on how the evolution in API access will impact the existing integrations in place.

Concerns about Security, scalability and vendor lock-in. High switching cost with infrastructure SaaS.

Business model - subscription vs. ad supported. Differentiation through service, experience and collaboration.

One of the panelist recounted how he sold his house from retouching images, creating flyer to final selling using only online tools. This reminded me of recent blog post by Phil Wainewright about how Appirio runs its business completely using on-demand integrated application infrastructure.

Panelists believed that SaaS margins will not be as high as SaaP and potentially declining, contrary to McKinsey's expectation of profitability improving as market grows.

Overall a great panel discussion event, thanks to great panelists and engaging audience.

Amazon Simple Storage Service (S3) is being considered a trailblazer and in an enviable position in infrastructure SaaS market. EMC/Unisys initiative shows product vendors caught off-guard with S3 growth.

How will SaaP vendors adopt to SaaS?

Cost vs. Benefit for Caching Appliances

2007-05-14T00:41:00.000-07:00

Gary Orenstein at Gear6 sent me preview material with his press release for CACHEfx appliance launch.

I guess writing nice things about vendors once in a while has some benefits. May be, I will be crawling into doghouse, by the time Gary finishes reading this post. :-( But, somebody got to ask the hard questions.

Gear6 Unveils Industry’s First Terabyte-Scale Caching Appliances to Accelerate Data Intensive Applications

CACHEfx appliances support a baseline of 250,000 I/O Operations per Second (IOPS), 16 Gigabits per second of throughput, microsecond response time and scale linearly to handle millions of IOPS. The Reflex OS™ virtualizes appliance memory into a scalable coherent cache pool, optimizes data delivery through parallel I/O channels, and provides robust intelligent cache services.

CACHEfx centralized storage caching solutions are available now starting at $400,000.

My jaw dropped after looking at starting price tag of $400,000. My first reaction was "Damn, this thing is expensive!" The performance stats like 250K IOPS and 16Gbps throughput are impressive but let's be realistic how many customers can afford to pay $1.60 an IOPS to speed up applications, beyond hedge funds and stocks/options traders. I am looking forward to a ROI/TCO justification in near future.

How big is the market for caching appliances anyway? I hazard to guess that $1.6 an IOPS eliminates most web companies with intensive data access performance as well as a large portion of HPC market.

When and where can caching appliance threaten the parallel and clustered storage solutions?

SaaS, Opportunity for Innovation

2007-05-13T17:03:00.000-07:00

I will be moderating a panel discussion SaaS, Opportunity for Innovation at Google campus Kirkland, Wednesday, May 16th. The panel discussion is part of May monthly meeting of our alumni organization IIT-PNW. I am looking forward to facilitating a great discussion among our panel guests and audience. Our esteemed panel hails from wide spectrum of SaaS (Software-as-a-Service) ecosystem.

Russ Arun, General Manager, Windows Live Communications, Microsoft. His current role includes Mail, Messenger, Manageability, Storage and back-end services for Windows Live.
Charlie Bell, Vice President, Utility Computing, Amazon. His current role includes the Elastic Compute Cloud (EC2) and the Simple Storage Service (S3).
Kevin Marcus, Chief Technology Officer, Intelius. His current role includes technology ecosystem of a SaaS offering.
Peeyush Ranjan, Engineering Manager, Google. His current role includes web search related projects with prime area of interest in building large scalable systems.

Would you like to attend the event? Please send an RSVP to IIT-PNW President, Mohan Venkataramana at mohan_13 [at] msvi [dot] org. Please mention this blog post in your message. The admission to this event may be limited due to the capacity restrictions of facility and IIT-PNW's primary responsibility to IIT graduates.

Would you like to be the "mystery" panelist? If you are opinionated, hail from SaaS ecosystem, don't belong to organizations already represented on the panel, send me an email with your contact info, interest, background and introduction.

It is going to be a very interactive event, slide-free, prop-free, whiteboard may be. All panel guests will have 5 minutes to introduce SaaS (What, Why, Where, Who, How) from their viewpoints followed by 40 - 60 minutes of audience Q&A.

Do you have a question on SaaS topic for our panel guests? Send me an email. Please indicate if you like your name and affiliation to be withheld. All responses will be posted on this blog.

Our last panel discussion Mobile Advertising - Technical Challenges and Business Opportunities moderated and blogged by Chetan Sharma had overwhelming response and very engaging audience. Unfortunately, being at Storage Networking World (SNW) in San Diego, I missed this great evening.

Look for a recap of the event sometime later in the week.

SVW: Storewiz. What do I like? Compression.

2007-05-09T22:10:00.001-07:00

There are three key factors (compression process, unpredictability of data reduction techniques, and allaying buying fear from startup) that are attractive for Storewiz. It doesn't appear that Storewiz is highlighting these factors publicly. I don't know, why? My cynical side cautions you to take these virtues with grain of salt. Of course, there is no doubt about the ultimate benefit with Storewiz appliance, the data footprint reduction on a storage device.

As you may realize (See previous posts, Storage Vendors to Watch: Storewiz. I and SVW: Storewiz. Q&A. and resulting comments), compression doesn't seem to get much love in storage industry with primary concerns being CPU utilization and performance impacts. How does Storewiz implement compression?

There is not enough information available from Storewiz on compression methodology and implementation. Most of the information below comes from Storewiz patent applications, specifically Method and System for Compression of Files for Storage and Operation on Compressed Files [US 2006/0184505 A1].

ABSTRACT. A method and system for creating, reading and writing compressed files for use with a file access storage. The compressed data of a raw file are packed into a plurality of compressed units and stored as compressed files. One or more corresponding compressed units may be read and/or updated with no need for restoring the entire file whilst maintaining de-fragmented structure of the compressed file.

The segments of an original file are sequentially compressed, by segment, into series of compression logical units (CLUs). The metadata for compressed section and corresponding CLUs are stored in a separate table.

Reading data stored in a compressed file requires identifying relevant compressed segment then CLUs belonging to that segment. Then, applicable CLUs are restored until all data that need to be read is restored.

Updating data stored in a compressed file follows the similar process as read. But, it involves a little more complexity as number of CLUs that need to be written after update can change from original number of CLUs restored.

Based on the patent document, the uniqueness in Storewiz compression implementation probably comes from:

Random access to data in compressed stored files
Operations on the compressed data without decompressing entire file
Compression/decompression operations transparent to users
User unawareness of the storage location of the compressed data

The compression in an appliance approach is the easiest, quickest and very flexible one to follow from initial product development and adoption perspective. Once, this method for block devices and communication matures, I don't see anything preventing from merging the functionality into storage and networking hardware as SBCs.

SVW: Storewiz. Q&A.

2007-05-07T23:25:00.000-07:00

Continuing from Storage Vendors to Watch: Storewiz. I

First the disclaimer for those with fertile imagination, I don't speak for Storewiz. Most information was obtained through public sources along with discussions at SNW with executives and others.

Let's address questions from Storagezilla, before I continue with my thoughts on Storewiz. He objected to the phrase "with better predictability than data de-duplication product."

What struck me about that is just like de-dup the data you're working with will dictate what savings you'll get. Image files or movies? Damn all or close to damn all. 1:1. … Databases or text files? Hell, you could get 5:1 compression, perhaps even more.

The data type is only one factor that impacts savings from data de-duplication. The others being "duplicity" of data in the dataset targeted for data de-duplication, "duplicity" of the dataset in stored data, targeted saving type, internal dedupe design and the implementation of data de-duplication solution in end-user environment.

Real life data de-duplication ratios vary a lot from 2:1 to 100:1. Data type on its own is not enough to be able to predict with certainty the achievable data reduction with data de-duplication. Beyond data type, data duplicity and the variations in duplicity over time is the main reason for a wide range in data reduction.

Why, when and where, "predictability" of data footprint reduction target matters more than the "highest achievable" data footprint reduction? Let's hear your thoughts first!

I've known about StoreWiz for a while now but I've always wondered where the FC/iSCSI compression boxes were?

Based on what I was told by Storewiz executives at SNW, the expected release is Q3/Q4, 2007. I am as usual skeptical of "Q3/Q4" claims like most people who have some experience dealing with any storage vendor. Most vendors say Q3/Q4 when asked in Q1/Q2 for time frame of next or new release. So, I wait too!

The sheer computational grunt required for such compression is an issue, …

I wrote in the last post, The strategy is similar to hardware compression in storage devices but with a twist that makes Storewiz implementation very resource efficient. Further explanation when I discuss three things that make Storewiz stand out in the data reduction market. As usual my opinion only.

All opinions expressed on this blog are my own, whether sink or swim. Your dissents, corrections and attempts to influence my opinions are always welcome.

Storage Vendors to Watch: Storewiz. I

2007-05-03T20:52:00.000-07:00

At SNW, I enjoyed briefings from two vendors the most. The enthusiasm of IBRIX executives for their product was contagious. And, simplicity of Storewiz product had me jump at the first opportunity to meet with company executives, JF Van Kerckhove and Jon Ash.

Storewiz product fits into two of the three themes, I observed at SNW, Global data reduction and Special-purpose appliances. It is a single-purpose appliance that helps you reduce the data footprint on storage devices with better predictability than what you get from data de-duplication products. Storewiz product provides real time on the fly compression, transparent to end user, easily added in front of the existing storage devices, and in some cases may improve performance.

What! that was my reaction when I first heard the performance improvement claims. Over the years, we all have been programmed to believe that compression slows things down and takes too many CPU cycles. My first reaction to performance improvement claims was no way! compression is an overhead, most probably slowing everything down. What is the first question that pops in to your mind when you hear compression?

The compression/decompression activity is performed by CPUs in Storewiz appliance eliminating the need to run compression process on storage devices or hosts. The strategy is similar to hardware compression in storage devices but with a twist (explained later in another post) that makes Storewiz implementation very resource efficient.

Storewiz execs also claimed that their appliance doesn't increase latency by more than 50 - 100 microseconds. I believe as long as the latency caused by compression process doesn't exceed the net decrease in time to write "compressed" data to the disks, you could potentially see the performance improvements during write operations. Same is true for reading the data from disk and decompression process which is also enhanced by read cache. Storewiz patented resource efficient compression implementation also reduces the need to compress/decompress too much data.

Checkout the image below showing NetApp read/write performance and CPU utilization with and without Storewiz appliance listed in one of their case study.

To be continued …

Storage Vendors to Watch: Cleversafe

2007-04-30T21:55:00.001-07:00

The second company with buzz at SNW, but not present at the show, wasn't a surprise unlike Gear6, a storage vendor to watch, in my opinion. As Clark wrote earlier, Cleversafe approach of security through obscurity was being considered a shift from traditional approach of encryption where encryption keys are single point of failure for data that need to be stored for a reasonable length of time. As previously mentioned, Cleversafe is one of my favorite new company also.

Paraphrasing below from Cleversafe patent application [11/241,555], Digital Data Storage System, the concept is simply to provide security through information dispersal and integrity through replication and hashing.

A distributed storage system for storing slices of original data on multiple storage devices in one or more locations. The individual data slices on each storage device are unrecognizable and unusable except when combined with data slices from other storage devices. The data slices are selected by information dispersal algorithms so that even if there is a failure of one or more storage devices, the original data can be reconstructed.

The ah ha! moment for me was when an end-user in a session at SNW asked a speaker for the opinion on Cleversafe grid strategy. Earlier, the same end-user had pointed out to me that he has been evaluating solutions based on clustered and distributed file systems. It also reminded me of last startup, I was involved with, where we were trying to utilize unused storage on untrusted and unreliable nodes within an enterprise. Our vision was more along the lines of FarSite than Cleversafe. We often encountered two questions that are primarily addressed by Cleversafe approach.

How will you make sure that data stored on untrusted nodes can not be accessed directly by users at that node?

How will you make sure that data stored on unreliable nodes is available even though one or more nodes may be offline.

I agree with Clark on the clever strategy adopted by Cleversafe to open source the code and look for revenue from service and support. I just can't visualize Cleversafe as a stand-alone product in an enterprise, more like a component of a larger grid based storage service or solution:

Leverage the inherent data protection available with distributed storage. Why should data first be pushed to a central location in the name of physical consolidation and then pushed out to duplicate in the name of business continuity?

Leverage the performance scalability with simultaneous transfer from multiple nodes. Why should data be stored on one node and restricted by the bandwidth and performance available at one node when it can be striped across multiple nodes to enable simultaneous transfer?

Trend of location proximity of data with user. Why should data be anchored at one place when user is becoming more mobile?

What else left to say?

2007-04-25T22:09:00.000-07:00

Storage Vendors to Watch: Gear6

2007-04-24T22:03:00.000-07:00

Last October, Gear6 came to my attention. But it didn't really hit home until a well-informed end-user talked to me about them at SNW. He was very excited about Gear6 product to address NFS performance issue with transaction databases. Based on what I understand, the product is a caching appliance, conceptually very easy to understand.

Excerpts from Gear6 website:

… keeping frequently accessed data in a very large central memory pool … This enables high performance data access by avoiding time-consuming disk operations and accelerates applications due to dramatically decreased response times and increased data throughput.

This innovative approach complements existing NAS/NFS deployments and installs transparently in the data center without requiring changes to current applications or infrastructure.

Actually, it was quite amusing at the conference. Most probably, I steered few end-users to Gear6 by suggesting to check out G6 caching appliance. These end users told me that they are using Oracle databases with NFS and performance being one of their pain points. I found three simple questions that can quickly tell whether someone may want to investigate G6 product.

Are you using transaction databases?

Do you use NFS mounts?

Do you have performance issues?

G6 product seems to be one of those products that require 10 minutes for presentation, 20 minutes for answering follow-up questions, 30 minutes for demo and then the question When do you want a unit for evaluation? Following is a clean version of a picture with G6, we drew at SNW. Is it feasible?
I will categorize Gear6 caching appliance as product that does only one thing but does it very well. With singular focus on NFS performance, Gear6 also made a good choice to attend Collaborate 07 an Oracle Community Event.

BTW, you may want to hop over to Thoughtput blog maintained by Gary Orenstein at Gear6 for more caching related information. Also check out these presentations from Gear6.

Share your thoughts on Gear6 and its caching appliance approach.

Storage Vendors to Watch: View from SNW

2007-04-23T22:25:00.000-07:00

Background

I had the pleasure of conversing with numerous end users, small to large, during three days of SNW conference. Typically, I found end-users to be more willing to discuss different topics and express opinions. I am not sure why vendors rarely can have an open and fun conversation. They tend to become clam.

Most end-users request anonymity and rightly so considering the prior troubles of one end user, I heard at SNW. An unscrupulous storage journalist published an overheard conversation with name of the end user, without permission, resulting in legal and HR issues for this person. Anyway, no name of end users on my blog. I met some great people at SNW with whom I hope to stay in touch for a long time to come.

I received vendor briefing from IBRIX, Axeda, Storewiz, Njini, Asempra, STORServer, VMware and Falconstor. Thank you to the executives from these companies to come and talk to me. I hope to provide digital ink in some form, favorable or unfavorable, to them in near future.

I also had interesting conversations for couple of hours with guys involved in M&A (Merger & Acquisitions) scouting storage companies at SNW. The experience with raising private equity for my last startup made discussing the prospects of various storage companies a fun exercise.

Most information in Storage Vendors to Watch series comes from my conversations at SNW.

Companies that were not present at SNW

Interestingly to start this series, I want to talk about two companies that were not present at SNW but were part of conversation at SNW. The reason, I am excited about these companies because all the initial information came from end-users currently investigating or evaluating their products. Hopefully, the end-users were not plugged in to SNW by these companies to disseminate positive information.

To be continued …

Themes of SNW Spring 2007

2007-04-22T21:15:00.000-07:00

It was a great experience to attend Storage Networking World in San Diego. Thank you, Bill Wrinn and ComputerWorld for allowing me to attend the conference with media credentials. Hopefully, blogger experiment was as interesting to organizers as to me.

Overall, I felt there were three main themes to the show:

Global data reduction: It was apparent at SNW that users are starting to look at products that can reduce the data footprint on all type of storage across enterprise. There is strong interest in de-duplication, compression or any other single instancing method. End-user Town Hall meeting is an exclusive affair. Data reduction may have been one of the topics brought up in the meeting!

On the topic of End-user Town Hall meeting, are you guys really think that taking your grievances from such closed meetings to vendors through SNIA helps? Security weaknesses in products didn't get addressed by vendors until they were made public.

Special-purpose appliances: End-users seem to be more interested in products that can do one thing very well than try to solve all their problems half-ass. Whether they are unification, performance, security or vertical-specific solutions, end-users didn't seem too concerned about having special-purpose appliances in their data center. Introducing delays in data path is still a concern but it is no longer a hurdle as long as a major pain point is resolved for the customer.

Another related mindset change was end-users willingness to buy from startups. In my discussions with end-users except for few very conservative shops, it wasn't uncommon for their data centers to already have or under evaluation products from startups. This is a welcoming change from couple of years ago when end-users were only willing to buy from large established vendors. Such conservative attitudes do nothing more than slowing innovation and creating barriers for real solutions to alleviate the pain points.

Clustered and grid storage: This is a shift from earlier trend of scale-up. No longer end-users are demanding a larger capacity and higher performing version of a product. Instead they are looking for product that can scale out as their requirements change without a need for replacing existing solution. Performance scalability through clustered storage is no longer confined to high performance computing (HPC) market. It started making significant inroads with vertical-specific applications and now moving in to main stream for resource intensive applications like data mining.

The success of Google with scale-out architecture using commodity hardware is also making end-users wonder why can't they leverage the same in their environment leading to increased interest in grid storage that can adapt to changing workloads, dynamic environment and provide single view of storage.

There could be another theme with data classification at SNW. But I didn't see anything that stood out, same old policy-driven stuff, nothing intelligent, with limited end-user interest.

Of course, you may disagree with the above themes so let's hear your thoughts. An IBM person with whom I shared cab to the airport thought the main theme was Storage Virtualization.

Go figure!

Readers Sentiment on Storage Blogging

2007-04-19T23:15:00.000-07:00

Mark wrote a retort to ex-bloggers who quit because of lack of readers, as I mentioned in my previous blog post after talking to few ex-bloggers at SNW.

In my opinion, most storage ex-bloggers tried to write for somebody else rather than for themselves. What they failed to do was to treat the readers, however few, as peers and build relationship. They just wanted to sell or show off something to their readers.

I firmly believe page hits, clicks, links and the number of readers don't count for much. What matters most is the level of interaction with readers, not only online through blog comments but also offline through emails, phone calls and face to face encounters. Differences of opinion doesn't matter, take solace in the fact that at least readers felt comfortable enough to share their opinions and express differences.

Even though directed at companies, The Cluetrain Manifesto gives advise that every storage blogger should heed:

Can you put yourself out there: say what you think in your voice, present who you really are, show what you really care about? Do you have any genuine passion to share? Can you deal with such honesty? Such exposure?

The reader sentiment, in my small sampling at SNW, about storage bloggers was very consistent. The readers expressed liking storage bloggers with personality and have one or more qualities of being opinionated, straight-shooters, analytical, passionate and even somewhat of nut-case. The blog readers in storage industry and in end-user storage community subscribe to most known storage blogs but they related to very few of them.

Actually, no corporate storage blog made the grade on building relationship with the readers, especially the readers from end-user community. The reasons for skepticism were quite varied like only covering topics aligned with the interest of their companies to when do these supposedly busy executives find time to write blog posts. Most readers wouldn't be able to differentiate most blog posts from the company marketing collateral if both were presented in same format. Some even assumed that topics and majority of content for corporate storage blogs may be generated by ghostwriters and marketing groups.

The data storage professionals who work in end-user environment also felt uncomfortable blogging about storage. The reluctance was particularly strong with ones who are involved in evaluation and design of storage solutions. They felt that the topics of interest to them like architecture, design, performance and problems are typically protected by vendors confidentiality agreement as well as fear of potential reprisal from storage vendors and their own management. My suggestion to end-users was to blog anonymously, as guest writer on other storage blogs or just feed the sources they trust, whether analysts, journalists or bloggers doesn't matter.

Hearing end-users at SNW unsolicited praising IBRIX, Isilon, 3Par, Gear6, and Acopia while thrashing products like SVC from established vendors, my advise to storage startups is to start blogging and also facilitate end-users to blog that are testing and using your product.

SNW Day Two 10:00am - Bloggers Evening Recap

2007-04-17T10:08:00.000-07:00

SNW Day One was really busy from early morning to late in the night. For me, the highlight of SNW Day One was the Bloggers Evening. The turn out at the evening was better than expected.

I sent out a reminder early afternoon to meet outside the hotel lobby bar at 6:00pm. Bruce Moxon, Blake Golliher, Clark Hodge, Claude Lorenson, SW Worth (both starting a Microsoft Stoarge blog) and myself initially met outside the lobby bar. We talked for half-an-hour standing outside the bar about blogging and storage.

Blogging concerns expressed during discussions were about making sure the information posted on the blog doesn't run afoul with employers as well as respecting the NDAs and confidential information of employer and suppliers. As we had a mix of storage bloggers representing vendors, services and end user, the blogging concerns were very varied. Past couple of days, I met quite a few ex-bloggers who were quickly discouraged not by their employers or any external entities but from absence of readers resulting in loss of blogging enthusiasm.

Unfortunately, we missed Marc Farley as I didn't get a chance to check email before Bloggers Evening and he had replied to my reminder mentioning being delayed for few minutes.

In the end, Bruce, Blake, Clark and myself went out for dinner to a nearby Italian restaurant. We had spirited discussion mostly about storage and blogging. We talked about how posts from opinionated and analytical bloggers make for a better read than people who just want to write nice things, marketing spiel or creating a very polished post. We also got in to discussion about our favorite upcoming storage companies, favorite conferences and events, favorite technologies and how changing storage landscape may impact few existing storage players.

Afterward, Blake and I also got in to discussion at hotel bar about social networking and its migration in to enterprise, our careers, interests, storage landscape and what's next for us.

Overall, Bloggers Evening turned out to be a great get-together. I echo the statement expressed by Bruce and others.

When is the next bloggers get-together?

I hope next Bloggers Evening is sooner than later.

SNW Day One 9:00am

2007-04-16T09:00:00.000-07:00

SNIA introduces new Executive Director

Last night, I attended SNIA event at SNW to introduce SNIA's new Executive Director Leo Leger. I got to meet some new faces and some old.

It was great to run into an old acquaintance, Laurence Whittaker from my days at Toronto Storage Networking User Group (TSNUG). He is as usual very active with SNIA End User Council. We didn't get to talk about the EUC strategy planning over the weekend as he got dragged of by Wendy. Hopefully, I will run in to him again during the show and get some information on EUC plans for the next year.

I talked to Leo Leger, the new SNIA Executive Director. The talk was quick and short kind of reminded me of being at speed networking event! I also sat down with Arun and others from Patni Computers. They seem to be active in providing software services to storage companies.

I also met Scott Kipp from office of CTO at Brocade/McDATA. He wrote several books including Fibre Channel Advances and Broadband Entertainment: Digital Audio, Video and Gaming in Your Home. He shared challenges and discouraging results from his past attempts at blogging. He is again planning to start a blog. I invited him to Bloggers Evening and talked to other bloggers about their experiences and feedback on his challenges. Unfortunately, he is not able to join us for Bloggers Evening due to conflict with SNW Speakers Dinner.

The highlight of the evening was talking to Vincent Franceschini. Even though, he doesn't post regularly, I always enjoyed reading Vincent's blog as he focuses on emerging technologies at HDS. Personally being interested in technologies ready to cross the chasm from research to industry, we had lively discussion on the challenges and opportunities of Grid Storage and Service-oriented Architecture. It was surprising that he had thoughts on the role of memory prediction in grid storage.

Talking to Vincent felt like sitting in a room brainstorming ideas. He is definitely passionate and opinionated about emerging technologies, the traits I admire personally. He mentioned enjoying blogging but also explained time constraints with current responsibilities both at HDS and SNIA as well the challenges of talking about future trends and emerging technologies that may be construed as pie-in-the-sky by some. Hopefully, he will take me up on my offer to work with him on blogging about emerging technologies as well as with Grid Storage initiatives.

SNW Solutions Lab

Last night, I also walked around SNW Solutions Lab where people were working hard to make everything operate properly. It reminded me of my involvement with SNIA SNW Interoperability Lab six years ago. I never got to see the fruits of the labor that time but from the planning chaos of 2001, this Solutions Lab seems lot more organized.

SNW Day Zero 6:50pm

2007-04-15T18:50:00.000-07:00

Storage can't compete with Aircraft Carrier

I arrived in sunny San Diego at 3:00pm to attend Storage Networking World Spring 2007. On my way from airport to Manchester Grand Hyatt hotel, the site of SNW, I passed by lots of boats and a huge Aircraft Carrier. And I couldn't resist the allure of aircraft carrier which overshadowed anything storage had to offer. So the first thing I did after checking in to hotel was to take a walk to fascinating USS Midway and the USS San Diego monument.

SNW before the Opening

After enjoying the sunny Aircraft Carrier, I checked out the different locations where SNW events will be held and took a peek in the expo hall before opening day. Looks so quiet but I am sure tomorrow, we wouldn’t be saying same.

Microsoft Water

While navigating around the huge equipments crates of various vendors, I ran in to none other than Darrell Kleckley and the Microsoft Storage team. I first met him when he was on SNIA Education Committee and now he is a Technical Evangelist at Microsoft.

I hope Microsoft sent cases of Windows Vista water to SNW. If Donald Trump can sell water then why not Bill Gates too! And if you listen to Paul Graham, water may be the next market Microsoft needs, to make itself feared again.

Bachelorette Party

It is just my luck that I get to stay in the room right across a bachelorette party room.

Can we meet at SNW?

2007-04-14T23:08:00.000-07:00

Sunday 4/15 noon, Leaving Seattle for San Diego.
Monday 4/16 evening, At Bloggers Evening.
Tuesday 4/17 from 10:00am to 7:00pm, In one-on-one meetings.
Wednesday 4/18 evening, Leave San Diego and return to Seattle.

Rest of the schedule is wide open. And, I would like to meet as many of my blog readers as possible.

If you are going to be at SNW and your schedule permits, let's meet. Just send me an email or call me with day, time and location where you would like to meet.

Bloggers in Demand at SNW

2007-04-10T21:09:00.000-07:00

Clarification: Bloggers Evening is Monday, April 16th at 6:00PM in Hotel Lobby. The JPR Cocktail event is Tuesday and totally separate from Bloggers Evening. I just wanted to clarify this as several people inquired.

I am amazed at the amount of attention being bestowed to storage bloggers at SNW by vendors, analysts and PR firms. I have received numerous emails and phone calls since my first post in which I mentioned plans to attend SNW as Storage Blogger.

Do you maintain a storage blog and would like to cover SNW events as blogger? Please send me an email or get in touch with Bill Winn at Topaz Partners who is managing media credentials for storage bloggers.

Several storage industry executives who presently don't blog requested to attend the Bloggers Evening. Clark Hodge and I discussed the issue of non-bloggers attending Bloggers Evening. We have decided to open up the Bloggers Evening to everyone.

Scott Kline from JPR Communications wrote a comment to my previous post and also sent me an email with official invitation for all Bloggers to JPR Cocktail Party.

I also want to have all of the Bloggers and associates of the Bloggers join us for cocktails and appetizers on Tuesday night from 7-9pm at the main lobby bar. I have attached the official invitation for you and whomever else wants to have some free drinks and food with all of the editors, analysts and companies.

Bloggers Evening at SNW

2007-04-08T09:48:00.000-07:00

Response exceeding Expectations

Response to my call for organizing an evening with Bloggers at SNW has exceeded my expectations. Mario Apicella aptly summed up my feeling about this initiative.

Only two years ago you could have counted storage bloggers without taking your socks off and now we can put together a small crowd.

Initially, I thought this evening may turn out to be party of one or two. But after a week of responses, I wish we should have tried organizing bloggers evening earlier.

In addition to last blog post, Say Hello at SNW, I reached out to storage bloggers through comments on their blog and email messages. I will continue to reach out to more bloggers next week. If you don't hear from me, it may just indicate that I am not aware of your blog. Please don't assume that you are not invited, just send me an email, leave a comment or pick up the phone and call me. All bloggers are welcome to Bloggers evening.

Thank you Storagezilla, Tony Pearson, Josh Maher, and Mario Apicella for spreading the word through post on your blogs.

A Party of Six

As of Saturday night, Bloggers evening is no longer a party of one but SIX. Following bloggers have expressed interest in attending Bloggers evening.

Clark Hodge, Storageswitched! Blog
Claude Lorenson, starting a storage blog at Microsoft
Marc Farley, Equallogic Blog
Tony Pearson, IBM Inside System Storage Blog
Jon Toigo, Drunken Data Blog

Even few bloggers, who are not attending Bloggers evening, mentioned the reason being absent all together at SNW. They also have very encouraging words for Bloggers evening.

"I'm afraid I can't make it to this SNW. … I love the idea of a bloggers get-together." Dave Hitz, Netapp

"it's a great idea but I am not going to SNW this time and this is one more reason to regret it :>)" Mario Apicella, Infoworld

"It is a great idea and thanks for the invitation!" Mike Linnett, Zerowait

When and Where

I am not familiar with SNW host Hotel, surrounding area and attendees interest in evening events at SNW. I am tentatively proposing we meet Monday, April 16th at 6:00PM in the lobby of Hotel Manchester Grand Hyatt and head out to a bar/restaurant at hotel or a location close by. Any alternate suggestions are very welcome.

Agenda

There is no specific agenda for Bloggers evening. Most likely, our discussion will revolve around storage blogging and data storage industry. If you would like to discuss any specific topic, please leave a comment, send me a message or just raise your topic at Bloggers evening.

Neither Fee Nor Free

There is no fee for attending Bloggers Evening, just bring your passion for blogging and data storage. But do bring your credit card, cash, food stamps, guns or any other method you use to pay for your own drinks and food. Unfortunately, there is no financial sponsor to cover the cost at Bloggers Evening.

Even though, Jeremiah will not be at Bloggers Evening, he has offered to buy us some drinks. Thank you! for the offer, Jeremiah. Anyone from PodTech is welcome to join us at Bloggers Evening. Anybody else who wants us to get drunk and stuffed, you are welcome to join us and pick up the tab.

Say Hello at SNW

2007-04-01T23:08:00.000-07:00

I am planning to attend Storage Networking World (SNW), April 16 - 19, 2007 in San Diego. I will be there as storage blogger and plan to blog often about interesting stuff from the conference floor, thanks to SNW organizers. Hopefully, I will meet lots of people from my favorite storage startups Cleversafe, Storewiz and InPhase, other storage vendors and users.

I am looking forward to seeing fellow storage bloggers, familiar faces from the past and some strangers who will become familiar in the future. If you will be at SNW and want to get together, leave a comment, send an email, call 206-390-4580 and give me a date and time when you would like to meet.

An Evening with Bloggers

I also would like to bring together all bloggers attending SNW for one evening. Please get in touch with me, if:

You are a blogger attending SNW and want to meet face-to-face with fellow bloggers.

You are interested in helping with organizing a bloggers evening at SNW.

You are interested in discussions with bloggers.

Suggestions for SNW Attendees

Unlike fellow bloggers, Jeremiah Owyang, Chuck Hollis, Dave Hitz, and Hu Yoshida, I attend very few conferences during a year and like to leverage such opportunities to the maximum.

What were your experiences at SNW in the past? What looks interesting this year? What suggestions would you offer to SNW attendees?

Use comments below to share your thoughts.

Data Deluge - Storage software need to step up!

2007-03-25T22:00:00.000-07:00

From EMC sponsored IDC study, The Expanding Digital Universe: A Forecast of Worldwide Information Growth Through 2010 [PDF].

In 2006, the amount of digital information created, captured, and replicated was 161 Exabyte (EB) = 161,000 Petabyte (PB) = 161,000,000 Terabyte (TB). Three million times the information ever written in books.

From storage perspective, 403 million hard drives of average 400GB capacity each without any RAID protection would have been required to store all the digital data, 93% of all hard drives produced in 2006.

In 2010, the information added annually to the digital universe will increase six folds to 988 EB.

To use same number (403 million) of hard drives in 2010 as in 2006, disk capacity will have to increase to average 2.4 TB.

The most telling statement in IDC study, I found, is:

In 2007 the amount of information created will surpass, for the first time, the storage capacity available.

This only goes to show that we just cannot rely on capacity of storage media to store all information created. Software has to step up to make sure that we can store all information created on the storage available to us. And that is why I believe technologies like data de-duplication and compression will finally become appealing in primary storage arena. See, my previous post Data De-duplication for Primary Storage.

Also, check out my new favorite storage startup Storewiz focusing on compression for primary storage.

Update: Canceled - Seattle Storage Event: Service-Oriented Architectures (SOA)

2007-03-11T21:28:00.000-07:00

Canceled

~~Checkout the local storage event organized by Puget Sound Storage Networking User Group (PS-SNUG) on Service-Oriented Architectures (SOA) later this month.~~

DATE: Thursday, March 29, 2007

TIME: 12:30 PM - 2:30 PM

LOCATION: SANZ Inc., 150 Nickerson Street, Suite 206, Seattle, WA 98109. Free parking available! Map

TOPIC: Service-Oriented Architectures (SOA): Designing Storage Environments to Match Business Needs

SPEAKER: Abbott Schindler, HP StorageWorks. Abbott Schindler has been in Storage development for 11 years; and a storage technologist with technical marketing experience for the past 15 years. He is currently involved with storage grids, virtualization and other leading edge strategies for HP. He has delivered extensive trainings and public presentations internationally, both internal and external to HP.

AGENDA: 12:30 PM: Pizza, Soda, and Networking. 1:00 PM: Presentation. 2:00 PM: Q&A. 2:30 PM: Conclusion and $50 Best Buy Gift Card Drawing

RSVP: To register, please visit SNUG website.

There is no charge to attend; however, we ask that you register for refreshment planning purposes. Meetings are open to anyone interested in discussing data storage in a vendor-neutral, education-focused environment.

Distributing Desperate Housewives to Ten Millions

2007-03-02T14:59:00.000-08:00

Now the title and image caught your attention, the big let down is this post has no housewives to offer! It is about distributing episodes of ABC Television show “Desperate Housewives” over Internet … or may be not even that!!

After my previous post P2P powered Devices … coming soon?, Newell Edmond co-founder of GridNetworks forwarded me an interesting paper on Video Internet. And this paper led me to March 2, 2006 column by Robert Cringely Peering into the Future: Why P2P is the Future of Media Distribution even if ISPs have yet to Figure that out.

"Desperate Housewives," in its puny 320-by-240 iTunes incarnation, occupies an average of 210 megabytes per episode. A full-resolution version would be larger still. In theory, it would be four times as big, but practically it would probably come in at double the size or 420 megabytes. But let's stick with the little iTunes version for this example.

Twenty million viewers, on average, watch "Desperate Housewives" each week in about 10 million U.S. households. That's 210 megabytes times 10 million downloads, or 2.1 petabytes of data to be downloaded per episode. Fortunately for the download business model, not everyone is trying to watch the show at the same time or in real time, so iTunes, in this example, has some time to do all those downloads. Let's give them three days. The question on the table is what size Internet pipe would it take to transfer 2.1 petabytes in 72 hours? I did the math, and it requires 64 gigabits-per-second, which would require an OC-768 fiber link and two OC-256s to fulfill.

Even though, Cringely was discussing bandwidth challenges of transferring one episode of Desperate Housewives, my mind wandered off to storage infrastructure side of the equation.

What type of storage infrastructure ecosystem will someone need to fulfill Ten million requests for distributing one episode of Desperate Housewives?

In my opinion, a storage infrastructure built around monolithic centralized storage most probably wouldn’t be practical. But this post is not about my opinion. It is about yours, so chime in with your thoughts on potential solution to this problem.

Show your design prowess or extol virtues of your favorite storage vendors with your storage ecosystem design. All responses are welcome.

Improving ROI of IT management

2007-02-26T22:21:00.000-08:00

Most IT departments are having difficulties in addressing IT infrastructure monitoring and management requirements due to

Limited resources
Increasing complexity of infrastructure
Growing number and type of devices in an environment, and
Demands for better usage, performance and uptime reporting

While attempts are made by industry to address these challenges through unified management platform, few years ago, I was told by an IT management vendor that over half of such products sit on the customer shelf due to integration complexity, customization requirements and wide variations in capabilities of various device modules provided by device vendors.

Incorporating device configuration and control functions to unified management applications is considered the major hurdle in adoption of such applications. But, IT administrators typically prefer to work with feature-rich device-specific configuration and control applications instead of using a generic unified management product. Most device vendors also provide limited functionality beyond monitoring and basic reporting to unified management platforms limiting value delivered through them. Rightly so, it also plays in to lock-in strategy and defense against out-of-sight out-of-mind perception for device vendors.

As more and more devices being installed in customer environment with email capabilities, real-time monitoring and auto-support from vendors, these management functionalities are becoming quite burdensome for the IT administrators, if not potential security risks. Just think aboutseveral dozen devices with frequent out-bound emails, monitoring pings and SNMP alerts in a data center and the management headache these auto-support functions can create.

So, how can the IT management vendors increase the value delivered to the customers with monitoring, reporting and analysis tools and without device configuration and control abilities as well as without creating another layer of management headache? As mentioned in my last post Understanding the Web 2.0 Trends, this was a topic of my conversation with Scot French, VP of Marketing at Klir Technologies, a local startup, for last few weeks.

I particularly liked the Klir solution to IT monitoring, reporting and analysis platform market with scalability of Software-as-a-Service (SaaS) and collective intelligence of “2.0”. I believe their approach brings the ease of use, perpetual upgrades, contextual content and community approach to IT management that is not available from enterprise IT management software packages.

Your opinions are welcome on IT monitoring, reporting and analysis market, SaaS approach and leveraging “2.0” to enhance ROI of IT management tools.

Understanding the Web 2.0 Trends

2007-02-25T19:11:00.001-08:00

Last week was quite active for me with lot of good discussions and brainstorming sessions in addition to spending an evening at Google office.

Challenges of delivering news through search
Continued obsolescence of current backup and recovery strategies
Improving ROI from infrastructure and storage monitoring initiatives
Web 2.0 and Enterprise 2.0 extending beyond applications to infrastructure

Note: Your thoughts are welcome on these topics.

The underlying idea loosely connecting all activities appears to be the “2.0” and its influence. And, these observations lead me back to reviewing What is Web 2.0, a 2005 article by Tim O’Reilly on understanding the Web 2.0 trends.

… the core competencies of Web 2.0 companies:
Services, not packaged software, with cost-effective scalability
Control over unique, hard-to-recreate data sources that get richer as more people use them
Trusting users as co-developers
Harnessing collective intelligence
Leveraging the long tail through customer self-service
Software above the level of a single device
Lightweight user interfaces, development models, and business models
Operations must become a core competency. So fundamental is the shift from software as artifact to software as service that the software will cease to perform unless it is maintained on a daily basis. It’s no accident that Google’s system administration, networking, and load balancing techniques are perhaps even more closely guarded secrets that their search algorithms. Google’s success at automating these processes is a key part of their cost advantage over competitors.

Lightweight business models are a natural concomitant of lightweight programming and lightweight connections. When commodity components are abundant, you can create value simply by assembling them in novel or effective ways.

How are you being pitched data de-duplication?

2007-02-20T21:30:00.001-08:00

Recently, I had interesting discussion with an IT professional about data de-duplication. It started with a very simple query on in-band and out-of-band data de-duplication that became a full-blown discussion and worth sharing on this blog.

Initially, I was taken back with the terms “in-band” and “out-of-band” being used to segment data de-duplication. I haven’t heard these terms since the days of storage virtualization. These terms were marketecture (nicer way to say marketing FUD) used by vendors trying to trip each other by pigeon-holing different solutions.

These terms leverage the long-held implicit connotation with in-band that anything inserted in the data path is not good for customer environment. And, true to the essence, this IT professional also assumed the same without really trying to understand various data de-duplication methods and their pros-and-cons. If in-band was so universally unacceptable, network switches would never have been introduced between server and storage. They are in data path, aren’t they?

Mute Debate: In-band vs. out-of-band

With storage virtualization, in-band referred to in data path and out-of-band referred to outside data path but typically in control path. With data de-duplication the boundaries between in-band and out-of-band are fuzzy unlike storage virtualization. In data de-duplication, all methods touch the data, the only difference is where and when this touch happens.

My suggestion was next time someone tells you that their solution is better because it is out-of-band, ask them how? And don’t accept an answer that doesn’t go deeper than “because it doesn’t sit in the data path.”

What are your pain points?

My next suggestion was not to lose sight of the pain points, you are trying to solve. If you are trying to decide how you want to go from A to B, your decision should start with mode of transportation not whether a car comes with V8 engine.

What are your pain points? Prioritize! There is no silver bullet.

Remote office / remote clients

Backup time window

Backup size footprint

Offsite backup

SMART not so smart in predicting disk drive failure

2007-02-19T13:27:00.000-08:00

Continuing from last blog post, Google report [PDF] also shares their analysis based on disk self-monitoring data and identifies important failure related SMART parameters.

The drives with scan errors are 10x more likely to fail that the drives with no scan errors. 30% of the drives fail within the 8 months after first scan error. The failure probability is higher within first month of first scan error occurring in newer drives and then plateaus. With older drives, failure probability rises with time.

The drives with reallocation count fail 3 – 6x more often than those with none. 15% of the drives fail within the 8 months after the first reallocation.

There is no definite correlation between failure rates and seek errors.

CRC errors are less indicative of drive failures than that of cables and connectors.

There is no significant correlation between failures and high power cycle counts for drives less than two years old. For drives 3 years and older, higher power cycle counts can increase the absolute failure rate by over 2%.

Predictive models based on scan errors, reallocation count, offline reallocation count and probational count couldn’t predict more than half of the failed drives.

We conclude that it is unlikely that SMART data alone can be effectively used to build models that predict failures of individual drives. SMART parameters still appear to be useful in reasoning about the aggregate reliability of large disk populations, which is still very important for logistics and supply-chain planning.

Glossary of Terms

Scan Errors – Large scan error counts can be indicative of surface defects, and therefore indicative of surface defects.

Reallocation Count – When the drive logic believes that a sector is damaged it can remap the faulty sector number to a new physical sector drawn from a pool of spares. Reallocation count reflects the number of times this has happened, and is seen as an indication of drive surface wear.

Offline Reallocation – Offline reallocation are defined as subset of the reallocation counts in which only reallocated sectors found during background scrubbing are counted. In other words, it should exclude sectors that are reallocated as a result of errors found during actual I/O operations.

Probational Count – Disk drives put suspect bad sectors “on probation” until they either fail permanently and are reallocated or continue to work without problems. Probational counts can be seen as a softer error indication.

Seek Errors – Seek errors occur when a disk drive fails to properly track a sector and needs to wait for another revolution to read or write from or to a sector.

CRC Errors – CRC errors are detected during data transmission between the physical media and the interface.

Google Findings of Disk Failure Rates and Implications

2007-02-18T13:28:00.000-08:00

Few months ago, I came to know that Google will be publishing a detailed report on disk drive failure rates in their environment. Prior to this report, best to my knowledge, there is very little information available from user perspective on this topic beyond limited work at Microsoft and Internet Archive.

Earlier, I thought of complimenting this report with a perspective from storage subsystem vendors. I was surprised to learn that either subsystem vendors don’t capture disk failure data effectively and in usable format or unwilling to share such data. And, there is very little published information on this topic from subsystem vendors.

In one case, a subsystem vendor recommended to contact disk drive manufacturers. Prior studies indicate that the actual drive replacement rate is 10 – 100x higher than failure rates published by disk drive manufacturers. Also, most likely disk drive manufacturers don’t have visibility in to deployment scenarios of failed drives for their data to be useful from the perspective of data centers and subsystem vendors. In the end, I decided to wait for Google report to come out to start this discussion.

Google Findings

I considered Google study to be unique because it looked at a very large sample of 100,000 disk drives. A summary of interesting results on age, manufacturers, read/write load and temperature from this study is listed below:

The failure rate varied from 1.7% for drives in their first year of operation to over 8.6% observed in their third year of operation.

Confirmation of the fact from prior studies that failure rates are highly correlated with drive models, manufacturers and vintages.

A complex correlation between high utilization, i.e. read/write load and higher failure rate instead of strong direct relationship as widely assumed.

First, only very young and very old age groups appear to show the expected behavior. After the first year, the AFR (annualized failure rate) of high utilization drives is at most moderately higher than that of low utilization drives. The three-year group in fact appears to have the opposite of the expected behavior, with low utilization drives have slightly higher failure rates than high utilization ones.

Surprising finding that lower temperatures are associated with higher failure rates and failures do not increase when the average temperature increases. The trend for higher failures with higher temperature is more pronounced for older drives.

Overall our experiments can confirm previously reported temperature effects only for the high end of our temperature range and especially for older drives. In the lower and middle temperature ranges, higher temperatures are not associated with higher failure rates.

Implications

In my opinion, the second finding reaffirms the subsystem vendors’ stance on not match-and-mix drives of different model and manufacturers.

The third finding has wider implications. First implication is on the higher possibility of failure of new drive during RAID rebuild even though this study explicitly didn’t measure very young age in hours after operation. Second implication is on extending time period for an old drive set to be kept intact and in operation after a new drive set has been phased in. Third implication is on the higher possibility of losing disks in long-term archive storage where read/write load may not be high enough and the challenges in keeping disk read/write load high enough.

P2P powered Devices … coming soon?

2007-02-17T10:48:00.000-08:00

Last night, I came across PBS's Robert Cringely post Appearances Can Be Deceiving: What's that 40-gig hard drive doing inside my Apple TV? on P2P technology incorporated in future generation of Apple TV.

If the Apple TV is plugged in it is turned on. Did you notice that? That means the hard drive will have at least the capability of running 24/7. Now envision a BitTorrent-like file distribution system that is controlled primarily by iTunes, rather than by you or me. A centrally controlled P2P system is VERY powerful because it allows for the pre-positioning of content.

His discussion of P2P powered devices is very interesting to me as few months ago, I presented similar thoughts in a series of posts on GridNetworks and what they can do to stand out in overcrowded P2P online video streaming market.

The second method, more likely to work for GridNetworks, is to pre-install or embed the player in to as many devices as possible, preferably the type of devices that are almost always on, almost always connected and publicly available to participate in content distribution without compromising owner-user experience.

They need to focus on embedding their player in to any device that has storage capacity and a network port. This may be the differentiation GN needs to stand-out in overcrowded P2P based online video streaming market.

Cringely's post is echoing same thoughts as I mentioned before about the success factors for GridNetworks, a startup in P2P space. Actually, he also gave a nice prop to GridNetworks in his post.

There are products like this already in operation, such as GridNetworks from Seattle or Mike Homer's Kontiki network, now part of VeriSign. It isn't rocket science, but to succeed, networks of this sort need lots of nodes, especially nodes that remain on 24/7.

Steve Rubel finds the idea interesting but doesn't think that Apple will have P2P based IPTV because of its focus on consumer products.

He may be looking at IPTV with similar business model as Cable TV. Instead, Apple IPTV most likely will be similar to combination of Apple iTunes and Apple TV with delivery between the two managed by someone like Akamai or GridNetworks.

Some may consider P2P for consumer applications only. In my opinion, P2P and Grid have wider applications in enterprise with open source projects like Cleversafe for dispersed storage.

Also see,

Challenges of High Quality Video Delivery
Success Factors for GridNetworks
Success Factors for GridNetworks … contd
Success Factors for GridNetworks … closure

Free Backup School, Seattle March 8, 2007

2007-02-12T23:30:00.000-08:00

Checkout a free Backup School in Seattle, March 8, 2007 organized by Storage Decisions and Storage Magazine. I received this notification through Puget Sound SNUG Mailing List.

For more details and the complete agenda, call Brian Digeronimo 508-621-5532 or register online at: http://searchStorage.com/r/0,,63031,00.htm?

"Backup School Hits the Road"

Where: Seattle, Washington
When: Thursday, March 8, 2007 (8:00 am - 4:00 pm)
Cost: FREE!
Who: This is a free seminar for IT professionals who are interested in improving backup efficiency and making restores more reliable.

Produced by: Storage Decisions and Storage Magazine

Your registration includes: Free breakfast and lunch, plus a chance to win one of two American Express gift certificates!

TOPICS INCLUDED:
• The top 10 ways people misconfigure their backup systems
• How to avoid those mistakes
• The pros and cons of the different disk-based backup targets
• An overview of CDP, data reduction backup and replication-based backup systems

Isilon Journey as New Entrant

2007-02-11T18:15:00.000-08:00

Last Friday, I had the opportunity to listen to Sujal Patel, founder of Isilon during Northwest Entrepreneur Network (NWEN) breakfast at Bellevue Harbor Club. He talked about the trials and tribulations last six years, he went through in bringing a new idea to mature market with Isilon.

It was an interesting presentation and hopefully, NWEN will post slides, notes or audio of his presentation online as they do with other breakfast meetings. Before discussing the four key factors of Isilon success, he talked about his background and Isilon timeline. This was on amazing story on its own.

Isilon Timeline
2001 - Series A funding $8.5 million
2003 - First customer, Kodak
2004 - NBC and Sports Illustrated using Isilon product for Athens Olympic
2005 - 100th customer
2006 - Initial Public Offering (IPO)

Establishing a storage startup during the height of dotcom crash. Isilon received Series A funding in 2001. He didn't talk about the journey from inception of idea to Series A funding, most probably another interesting talk on its own. But it was clear from his mention of 150+ meetings needed to raise Series A that it was not an easy path from idea to Series A.

Founders with no background in storage industry. He did have background in digital media delivery infrastructure. He started out solving the problem with digital media delivery instead of creating a storage solution.

Surviving when dozens of storage startups started by seasoned storage industry professionals failed. I think there were some inaccuracies in his list of storage startups RIP but the message was quite clear. Industry experience is not a corollary to a disruptive technology, solution or successful startup. It maybe an hindrance than tool to see a real customer problem and produce a solution, customers want.

His clustered storage product focused on solving a very small problem in the larger realm of storage ecosystem but an important problem to a niche digital media segment. Contrast that with technologies like iSCSI and CDP, the solutions still searching for a problem to solve, important enough to someone to pay for it.

Isilon, a case study in lessons of Clayton Christensen's The Innovator's Dilemma and Geoffrey Moore's Crossing the Chasm

~~More in future posts.~~ Update: John Cook posted a nice summary of Sujal's talk. As comments on his blog reflect, I also thought Sujal presentation was excellent, well-prepared and with right content for the right audience.

BTW, I asked him the last question about organizational changes taking place at Isilon as it is growing. The follow up question, I didn't get to ask would have been, how these changes will impact future innovations at Isilon.

One day, I would like to meet him as well as his product manager, Sam for whom he had very nice things to say.

Sights of Tokyo

2007-02-06T22:46:00.000-08:00

View of Peace Bell from cab on our way to hotel from Tokyo station.

View from the hotel in the morning.

View from Tokyo TV Tower

View from Tokyo TV Tower, looking down through plexiglass floor tile.

Just a bout of writer's block. :-(

Data De-duplication for Primary Storage

2007-02-01T23:35:00.000-08:00

Last night, while reading Steve Duplessie's long travelogue, I noticed a very interesting note referring to data de-duplication for primary storage.

We landed on time, and a car was nicely waiting to pick me up to bring me to my first meeting, with Data Domain … I chatted with a bunch of their smart folks theorizing about where other implementations of this technology could really affect change in the world, and found quite a few. What if you could get the performance attributes required by a high percentage of today's applications on a primary store that happened to get 40 to 1 compression rates? Imagine the economic advantages and the consolidation potential.

Again tonight, I came across another note referring to data de-duplication for primary storage in Jon Toigo's praise the PR agency post.

What de-duplication doesn’t address is the primary storage issue. The device used for primary storage is not inexpensive and has limited capacity. When you run out of space, you run out of space. You can manually delete and/or move to tape, but this is somewhat time intensive. Or, you can always make a storage vendor’s day by buying more and more hardware. What I believe will be a hot and extremely important technology in 2007 is data compression. The business case for data compression on primary storage is the same as the one used for de-duplication for backup. Compression can cut primary server data to a minimum of one-third its original size. It makes good business sense: by compressing data on primary storage devices you need less hardware less resources to manage, lower power consumption (and power consumption is a big deal), etc., all without a performance hit.

Hmm … it may be just my imagination but seems like Data Domain may be planning to enter primary storage market with a data de-duplication product and JPR may be doing guerrilla PR for them.

Anyway, I am glad to see that I wasn't the only one theorizing the extension of data de-duplication beyond backup to other storage sub-segments, last year. See Where are you being de-duped?.

Some of the near-term applications are going to be in the area of backup, archive, wide area data transfer, data caching, primary storage and enterprise data storage grids. In the end, data de-duplication can be applied anywhere where cost of resources freed by eliminating repeating patterns exceed the cost of resources required to remove repeating patterns.

Online Backup Services - What's Next?

2007-01-30T21:47:00.000-08:00

Earlier this month, Jeremiah and I were having debate about prospects of online data storage. At that time, I invited people who may be involved in online data storage services to contact me. I am continuing to receive responses from various industry professionals on this topic.

One of the person responded was Damien Stevens, CEO of Servosity, an On Demand Backup company. Using emails and phone call, we are having a great conversation around online backup services, where it is today and where it needs to be. In my usual direct manner, I started with a very specific question in an email.

AG: How is your approach unique compare to other online backup companies?

DS: Right now we're winning customers over the competition because we've made it easier to understand, we don't charge for backup agents or extra computers, and we are not overpriced like some of the higher end competition (redacted name of competitors).

In last four years, I have heard the same pitch combining ease of use and cost too many times from online backup providers and, I have seen dozens of them come and go in the same time frame. So I wasn't very excited.

Then, he piqued my interest wanting to discuss the technologies that may enhance the value of online backup services. The demand for these technologies is not something that can be delivered on overnight. Customers are not asking the benefits of these technologies right away either. But if customers were to ask for them in 12 - 24 months, it needs to go on the service/product development roadmap soon.

Since then, our conversation has been very lively and wide-ranging. The discussions revolved around potential customer requirements of versioning, lowering network bandwidth utilization, making restores easier, and better backup/restore job management. We also discussed wide variety of technologies like CDP, data de-duplication, snapshots, replication, archiving, monitoring and management.

In the end, I decided to leverage the power of blog by writing this post and putting forward questions to my readers.

What is your opinion and experience with online backup services?

As a customer and user, what would you like to see in an online backup solution in next 12 - 24 months?

As a service provider, storage vendor, or someone familiar with various storage technologies, what technology would you recommend to be included within next 12 - 24 months?

BTW, if you have any experiences with Servosity On Deman Backup service, I would like to hear about them.

Bloggers are like sheep herd

2007-01-28T23:42:00.000-08:00

I was absent from blogosphere for past two weeks, not intentionally. Blogging took a back seat while I managed few other projects in my life.

At work, by pure chance, I landed up performing the first field installation of new data de-duplication product. The pressure to not screw-up was great considering the high visibility internally. At home, I made a move from Eastside to Seattle that disrupted blogging routine. Also, as I am settling down in Northwest, moonlighting opportunities are growing resulting in thoughts about resurrecting a moonlighting venture.

Blogosphere hasn't been very exciting either. Most blog posts are starting to look very similar and repetitive. Everyone seems to hop on to latest news. Most posts are nothing more than hashed, rehashed or paraphrased versions of original story. It was amusing to see the coverage, blogger after blogger gave to Intel 45nm processor and the YouTube announcement of paying content creators.

Another annoying trend is promotion of audio and video podcasts through blog posts. I rarely listen or watch podcasts. Why? Because it takes too long to figure out quality and relevancy of podcast compared to scanning a text post in few seconds. And, more often than not, podcasts are nothing more than someone droning on and on about nothing .... It wouldn't be a bad idea for audio and video bloggers to start posting transcripts.

I am subscribed to almost 200 blog feeds through Google Reader. Still, there seems to be lack of original and quality content to read on blogs. Finally, I started axing the blog feeds that didn't have enough original content and thoughts put in to the posts. Quality is trumpeting over quantity in deciding who stays.

Old fashioned print publications and broadcasters seem to be offering better quality content like Z-RAM coverage in IEEE Spectrum. There may be a business opportunity for print publications or another startup, read all bloggers opinions on an issue or news, distill, summarize and print them.

Bloggers are expressing more and more what corporate marketing and public relations groups are feeding them than speaking their own mind and opinions. Bloggers are acting like sheep herd instead of sheep dogs.

Two Myths about iSCSI Adoption

2007-01-15T11:22:00.000-08:00

… Continuation of my opinion from previous post to the expanding iSCSI debate in the storage blogosphere.

Check out opinions of Tony Asaro and Steve Duplessie on debate between Dave Hitz & Chuck Hollis. Beth Pariseau also wrote a nice summary of the debate. Chris Evans is taking on in-depth investigation of iSCSI. I am looking forward to reading about his experiments and analysis.

Since my last post, I received few emails and phone calls with skepticism, typically bestowed on me, after writing an opinion that doesn’t align with storage people with fancy titles and so-called experts. Without going into specific details, I worked with iSCSI time-to-time for over six years in variety of roles on variety of iSCSI projects.

And, if you are just starting to understand the issues with iSCSI adoption explained in Chuck’s post, you have lot to catch up.

Controversial Myths

Chuck expressed some good thoughts about the future trends and driving factors for iSCSI adoption.

In my experience, some of the driving factors may be just red herrings. These are the myths that originate from wishful thinking of people promoting iSCSI instead of realities of the iSCSI adoption impediments in customer environment.

First, I think you’ll see more of the same: iSCSI in smaller, greenfield SAN builds where FC isn’t entrenched yet. But from small acorns mighty oak trees grow.

Myth #1 The smaller greenfield iSCSI environment will grow without any challenge.

This myth is created when people take the “switching costs can outweigh any benefit of a new technology” lesson too literally. They forget that as long as a technology is not fully entrenched, it is not very difficult to get people to switch. And that is the challenge for smaller Greenfield iSCSI installs. They are not fully entrenched yet thus not difficult to switch to something else like FC.

Two key points to ponder are:

Why greenfield iSCSI environment came about in the first place?

What happens when these small installs are considered for expansion?

Going out a bit farther, I believe that further cost reductions in 10Gb ethernet technology will encourage people to take another look at an alternative to FC.

Myth #2 10Gb Ethernet will save the day for iSCSI.

This myth is created when people assume that major hurdle to iSCSI adoption is speed and feeds. If history is telling, a new technology doesn’t replace incumbents just because of incrementally higher performance. Either new technology has to blow the incumbents out of water with several folds increase in performance or offer a value proposition that is not present in incumbents.

Just consider this, nobody would care about data de-duplication if it was offering 2x increase over conventional compression, but a 20x increase gets everyone’s attention. From another perspective, today you will be considered underdog with a Virtual Tape Library (VTL) solution that doesn’t have de-dupe even though VTL with de-dupe is slightly slower performing due to extra de-dupe processing overhead.

Another underlying assumption of this myth is that performance of everything else connected to 10GbE will remain the same as with current 1GbE and data transfer requirements will not grow. Fat chance!

Before 10Gb Ethernet can be considered for underlying iSCSI infrastructure, it also needs to be fully entrenched in customer environment as preferred networking infrastructure technology. There are too many infrastructure and political challenges to kill any 10GbE adoption for iSCSI environment without the Ethernet owners having the first crack at it.

The above observations are based on my past experiences with few expansion projects for small iSCSI installs. One of the project even considered 10GbE. In the end, it was too easy to replace iSCSI install with FC-SAN than expand them.

Controversial Positioning

I close this post with two controversial iSCSI positioning statements that I believe are impeding it’s adoption. No explanation provided at this time and you are welcome to chime in, if you desire.

Position #1 iSCSI is a storage interconnect leveraging Ethernet.

Posiiton #2 iSCSI is a replacement for or complementary with FC.

Year of iSCSI or Languishing iSCSI

2007-01-11T00:42:00.000-08:00

There is interesting iSCSI debate brewing between Chuck Hollis of EMC and Dave Hitz of NetApp.

I got my first taste of iSCSI when I laid my hands on newly released Cisco 5420 iSCSI router. The iSCSI standard was barely at draft 0.8.

Initially, I was excited about iSCSI and even got involved in SNIA iSCSI and IP Storage initiatives. But the way iSCSI was being positioned in the market, it didn’t take long to realize that iSCSI will play second-fiddle to Fibre Channel (FC) for a very long time.

T/M/C vs. Company

Dave compares first four years of iSCSI revenue with that of NetApp to convey that it feels like the “Year of iSCSI.”

In my opinion, Dave is thinking about it wrong because he is comparing oranges to apple. iSCSI is a technology/market/category (T/M/C) and NetApp is a company. A comparison between them makes no sense.

First four years of iSCSI revenue are divvied up between several iSCSI vendors with no single vendor dominating the iSCSI category. In contrast, most NAS revenue, not all, went to one company, NetApp. In sixth year of NAS, NetApp was equated to NAS. Can you tell me which company is equated to iSCSI in its sixth year? Nobody.

Whom do you think of when someone mentions NAS, server virtualization, data de-duplication, Fibre Channel, network routers, desktop operating system? Yes, whether it is NetApp, Vmware, Data Domain, Brocade, Cisco, Microsoft or countless others, there was one dominant vendor to born out of each of these “new” T/M/C. They were primarily responsible for defining the “new” T/M/C. And, this is the anathema for iSCSI in my opinion.

There is no single vendor that defines iSCSI category and until there is one, it will continue to languish.

There is another lesson in languishing iSCSI. A T/M/C languishes whenever it goes through standardization before any one vendor has a chance to define it, another successful strategy in the arsenal of established players for killing new T/M/C. This is why SNIA, primarily dominated by established vendors, keep pushing to standardize new T/M/C before they have chance to take root.

I am glad to see that SNIA didn’t get their claws on data de-duplication before Data Domain had a chance to establish roots

More to come ….

The Great Ideas of Dr. Mortimer Adler

2007-01-10T00:56:00.000-08:00

After reading my previous blog entry on Richard Feynman, one reader suggested picking up works of Dr. Mortimer Adler for my next non-storage related reading. Dr. Adler was a philosopher, professor, author and star of his own TV show.

After some research, I decided to get How to Think about The Great Ideas book from King County Library System. This book is a collection of transcripts from his TV show, The Great Ideas, aired in early 1950s. Guessing from how quickly my request for this book was fulfilled by library system, works of Richard Feynman seem to be more popular in Seattle region.

This week, I am focusing on reading this book instead of listening, reading and writing two cents in storage blogosphere. The Great Ideas book is very fascinating and some of the lessons are worth sharing on this blog.

Look for more from this book in coming days.

All the news about stormy conditions in Seattle region reminded me of this image of tree hanging on power cables. I captured this picture during last windstorm in Redmond.

Who has the best full disk encryption solution?

2007-01-04T23:14:00.000-08:00

Last week, a ZDNet blog post mentioned that US federal government is conducting a contest to find the best full disk encryption product. The winner will become the preferred federal solution to address data loss issues from stolen or missing computers.

The who's who of storage and encryption industry like Seagate, CipherOptics, Credant, Decru, NetApp, McAfee and PointSec attended the industry briefing. Click here for RFP information on Data at Rest (DAR) Encryption program.

The RFP DAR Technical Requirements provides high level criteria for a DAR solution.

FIPS compliant algorithms for encryption, hashing and signing
Encrypt data on removable storage media/devices
Single management console for Full Disk and File Encryption
File compression and encryption in a single step
Boot authentication

I wonder what are the different techniques for DAR encryption?

Online Data Storage Debate

2007-01-03T23:52:00.000-08:00

Check out friendly debate with Jeremiah about prospects and challenges of online data storage services. Actually, we first started this discussion during my visit to HDS in September.

Jungledisk connects you to Amazon S3

Why I think future Online Data Storage companies will Pay You to Upload Data

Surely, Online storage services such as Amazon's S3 storage service can be a useful service as secondary storage or data I want to access on the road or data I want to share with others or data that has low value to me. But as primary storage and for my valuable data, I am skeptical.

Can you trust anyone else to store and protect your data?

I will cover more on this topic in my future blog posts. Few weeks ago, I was asked about my thoughts on this segment by an acquaintance in financial sector. I am just gearing up for a detailed analysis of online data storage opportunities.

Are you involved in online data storage service segment? If you are, I would like to get your perspective for my analysis and future blog posts. Interested, you know how to contact me. >>---------->

I am wearing Montsuki and Hakama, traditional Japanese clothing for formal occasions, in this picture.

Gigantic Banana over Texas Sky

2006-12-28T21:48:00.000-08:00

This is funny! A project to put geostationary banana over Texas. Why didn't they pick a place in Canada? The project leader, Cesar Saez, is based in Montreal, Quebec.

I wasn't impressed with their reasons for picking Texas. They are just making mockery of an experiment that could have been great scientific and engineering adventure in addition to artistic expression.

IMO, the political undertone may alienate people from scientific community.

Success Factors for GridNetworks … closure

2006-12-28T20:05:00.000-08:00

Wrapping up my thoughts on prospects of GridNetworks. Continuation from previous post.

The second method, more likely to work for GridNetworks, is to pre-install or embed the player in to as many devices as possible, preferably the type of devices that are almost always on, almost always connected and publicly available to participate in content distribution without compromising owner-user experience.

One of the biggest challenge in my opinion is to get people to install a P2P based player on their computers. I personally started the installation process several times and backed out. The main concern was the impact of a P2P player on my work machine, both performance and security aspects. Pre-installed player and embedded appliances are a good method to overcome such concerns.

Another challenge is to have as many nodes public and participatory in content delivery as possible. Most computing nodes where users install GN player are more likely to be located behind a firewall/broadband router whether in the office or at home. This may result in a disproportionately higher number of nodes acting as freeriders without aiding delivery to others.

Newell claimed that their player is successfully running on Xbox. Embedding in Game consoles is definitely a step in the right direction. Some of the other devices, they may want to consider for embedding are:

Network routers - The ones normally used with broadband connections at home from companies such as Linksys and D-Link.
TV Set-top boxes - Comcast provided me a Motorola HD cable box with Ethernet port. May be GridNetworks can help set-top box manufacturers utilize this port for some meaningful purpose.
Helping NSLU open source community to install player on Linksys NSLU2 storage device.
Develop, partner and market NAS and Media PC devices with embedded player for consumer markets.

They need to focus on embedding their player in to any device that has storage capacity and a network port. This may be the differentiation GN needs to stand-out in overcrowded P2P based online video streaming market.

Good luck to Newell and his team at GridNetworks.

Communication using hand/body gestures at dinner table in Hokkaido, Japan. I knew no Japanese and others knew very little English.

Asempra Hiring while Others Firing!

2006-12-27T20:49:00.000-08:00

Last week, Eric Herzog sent me an email mentioning that he is looking for Systems Engineer, based in Sunnyvale and NYC, for Asempra. He is also searching for Director of Business Development and Alliances, Senior Technical Support Manager and Director of Product Marketing and Channel Marketing, all based in California. In his words,

All MUST have very, very SOLID storage subsystems knowledge but who know software as well as hardware.

If you are interested in these opportunities, apply through Asempra Career section. Don't forget to mention you came to know of these opportunities through this blog.

Please DON'T email me your resume unless you can write me a note explaining how your resume can make Eric dance in front of Asempra logo and send his dancing picture for me to post!

Excerpts from provided SE position overview:

The Regional SE position is the focal point of technical responsibility within the Sales organization for all pre-sales customer and partner engagements. The Systems Engineer is responsible for opportunity qualification, product demonstrations, technical presentations, on-site product installation, training, and, coordination with support and post-sales resources. Heavy regional travel is required - up to 70% on-site work with customers and partners.

The successful candidate will have a proven track record of working with and directing technical sales teams; The ability to influence and educate prospective customers and partners via demonstrations, technical resources or other methods to more effectively sell technical solutions; a deep understanding of the role of a systems engineer; strong analytical capabilities; a creative problem solver; work well in a dynamic team environment.

Ideal background would include expertise within Windows and Linux Environments, SQL, DR and Recovery Management, Storage Management; and the data protection market (backup, replication, CDP, volume management).

If you didn't like recent Steve's Rant and like Storagezilla wished Steve lump of coal, you may be a good candidate for Asempra.

Success Factors for GridNetworks ... contd.

2006-12-22T00:23:00.000-08:00

Continuing my thoughts on success factors for a video distribution infrastructure play like GridNetworks from previous post ...

As Michael Gersh commented in previous post, high quality video distribution will be viewer paid.

Who is going to collect payment from viewers? Will it be a content distribution infrastructure owner like Comcast or content distributor/aggregator like Netflix? Why is it important? IMO, it is the company in value chain that has most viewers captive benefits the most. And this is shown very clearly from some back of the envelope calculations for iTunes and Akamai.

Assuming weekly revenue of $10 million from analyst download estimates of $18.5 million songs per week, annual revenue of iTunes store, a content aggregator/distributor, is over $500 million+. Akamai, a distribution infrastructure provider to iTunes and with near monopoly in CDN, total revenues are barely in $400 million range.

With the success of iTunes, it is assumed that content aggregators/distributors are the ones who will be collecting payment from viewers. Distribution infrastructure owners like Akamai will be a service provider to iTunes for a fee.

Cost-side Success Factor

As GridNetworks (GN) will likely be paid by content distributors, it's profit-side success factor depend on number of content distributors using its delivery infrastructure and the revenue generated from each content distributor.

To attract paying content distributors and be a preferred delivery method, GridNetworks need to have the most expansive hybrid CDN P2P infrastructure. So the cost-side success factor for GN comes down to how quickly they can build 40 million quality nodes contributing to their delivery infrastructure and at what cost.

One method to achieve this goal is to freely distribute software for media sharing and playback. Once there are sufficient nodes established, harness those nodes and the brand recognition to make deals with content developers, owners and distributors. BBC deal with Azureus [pdf] will fall in to this category.

There are already enough high quality video content delivery startups trying to follow this route. Most with very little value differentiation originating primarily from the high profile and visible content deals. The success will belong to the ones with market/brand recognition, deep pockets and influence to make high-profile visible content deals.

Should GN follow the same path or there is another way to succeed? Chime in if you have any thoughts.

To be continued later ...

Housekeeping Notes

Recently, I noticed in Google Analytics stats that some traffic to my blog is coming from penny stock forums where link and text of my entries were posted. A caution note for readers from these forums: My posts are nothing more than personal rants and shouldn't be considered thoroughly researched analysis on prospects of any company, its stock, industry or market. Believe in my rants on your own perils.

This picture of a Pachinko in Tokyo seems quite appropriate after the above housekeeping note.

Success Factors for GridNetworks

2006-12-20T00:02:00.000-08:00

Note: For background information on my interest in GridNetworks, please read my previous posts:

Challenges of High Quality Video Delivery

GridNetworks, what's my Interest? Part one.

GridNetworks, what's my interest? Part two

Based on my observations of startup universe, I believe that every startup needs to have at minimum two success factors, one on profit side and one on cost side, to succeed.

Profit-side Success Factor

For GridNetworks (GN), the profit-side success factor is dependent on the number of distributors of high-quality content willing to pay for using GN infrastructure, irrespective of whether they are well-known content distributors with large content library like Disney or niche distributors leveraging long tail phenomena like Reeltime.

Some might argue advertising may be a viable option. I seriously doubt viability of an infrastructure intensive play like distributing high quality video solely based on advertising. Capturing $10 in advertising while paying $28 for delivery doesn't look like a viable business proposition to me. See, Comment on A-List Blog may result in high traffic and Internet Video Business Challenges.

A subscription based viewer model may become viable option for predictable delivery quality once the delivery infrastructure has achieved a critical mass of viewers, content, participatory nodes and content distributors. Most probably, only to be realized in next phase of GN.

To be continued ...

Honorable 'Whatis' Mention

2006-12-18T21:18:00.000-08:00

My blog is one of the seven blogs mentioned in Storage section of Our Favorite Technology Blogs at Whatis site of Techtarget. Thank you guys (and, gals!) of Techtarget for including this blog in to your list of favorite technology blogs. The others listed are Who's who of storage!

Dave Hitz
Jon Toigo
Hu Yoshida
Mark Lewis
Marc Farley
Robin Harris

The complete list of technology blogs published by Whatis provides wealth of information on regular basis. I just wish they had an OPML file of RSS feeds for these blogs. With OPML file, it will be a breeze for everyone to import all feeds in to Google Reader and Firefox Live Bookmarks.

Patent Search through Google - How Convenient!

2006-12-15T19:20:00.000-08:00

Google makes patent searching easy with Google Patent Search. See Google Launches Patent Search Beta, a TechNewsWorld article for more information.

Results for Avamar through Google Patent Search. Convenient and nifty! Hopefully, one day it will expand to include patents from other countries also instead of U.S. only.

View of Tokyo from Hotel room window

GridNetworks, what’s my interest? Part two

2006-12-13T23:03:00.000-08:00

Click here for the first part.

Third, GridNetworks piqued my interest with the use of word “Grid” in the company name. Grid is a powerful concept for utilizing idle resources and addressing infrastructure scalability. GridNetworks provides the extra scalability through P2P to fixed infrastructure of content delivery networks for superior user experience at the end of the last mile of Internet.

Few years ago, I researched a company for a client, Bycast and its StorageGRID product. During last bubble, Bycast was positioned as online video streaming startup. With its underlying technology, Bycast was successful in navigating turbulent times by repositioning in healthcare storage. Even though, Newell was quick to shoot down my comparison of GridNetworks with Bycast, I felt that the underlying technology of GridNetworks will find its ways in to other areas and applications too, not discounting its application in storage either.

Fourth, GridNetworks website listed Sujal Patel on its Advisory Board. Sujal is co-founder of Isilon Systems, a storage startup, soon to be public, and a veteran of RealNetworks. This piqued my interest to find out what connection Isilon or storage may have with GridNetworks.

In the end, it turned out to be nothing more than ex-RealNetwork links. Sujal is now acting more like angel investor. So, he doesn’t seem to be a good proxy anymore to find ties with Isilon or storage.

Fifth, the hints of P2P implementation in PowerGrid End User License Agreement piqued my interest as I saw a way to accomplish Distributed Storage Aggregation (DSA). And if you like me scan IEEE and ACM publications regularly, P2P way to utilize unused storage distributed across enterprise has started to become interesting.

Newell claimed about 50,000 nodes currently contributing on, average, 1GB of storage and 200kbps of bandwidth per node. His users are contributing almost 50TB of storage and 10Gbps bandwidth to deliver high quality video. When his vision of 40 million nodes comes to fruition, he may have access to 40 PB of storage without shelling out a single penny to any storage vendor.

The above reasons may explain what encouraged me to initiate contact with GridNetworks. One day, I expect to see the same technology being implemented with in the enterprise for distributed storage aggregation.

Excerpts from PowerGrid EULA:

"Segment" means a small block of encrypted Content data, typically under one megabyte.

Encrypted, managed peer-to-peer "grid" network architecture.

The Content Segments stored on your computer may be portions of files you have viewed, or may be copied to your computer by GridNetworks host systems for later sharing with other GridNetworks authorized users.

The Segments contained therein are only accessible through the GridNetworks PowerGrid application and cannot be individually viewed, altered or deleted through your computer's file management tools.

These Content Segments are periodically removed and replaced with other segments and shall never exceed the maximum capacity of the Data Store, established by you.

City view from Tokyo TV Tower. Mount Fuji is barely visible in the background.

GridNetworks, what’s my Interest? Part one.

2006-12-11T23:24:00.000-08:00

As I mentioned in my post, Challenges of High Quality Video Delivery, last week, I spent an hour with Newell Edmond, Co-founder and Amy, Project Manager talking about GridNetworks, its technology and business.

Contrary to what few readers thought and queried, neither I met them for a job nor they rented me to blog about them. I discovered GridNetworks through John Cook’s blog post Getting 'Goodfellas' on the grid. It piqued my interest for several reasons.

First, I have been thinking of ways to highlight startups working on interesting infrastructure solutions. Consumer Internet and Web 2.0 startups are topic of discussion by bunch of bloggers. But the infrastructure startups that enable them are largely ignored.

There was nothing better than a local infrastructure startup to initiate my coverage. Expose your infrastructure startup, you know whom to contact and how!

Second, even though my initial impression was that it is yet another startup trying online video delivery. Further research showed their aspirations are more substantial than just becoming a video destination.

And after meeting with Newell, I am convinced that they are actually a video delivery infrastructure play, combining content delivery network and peer to peer network technologies, than just a torrent player. And this is also confirmed by Michael Gersh, a VP at Reeltime, a GridNetworks customer, in his email.

We are currently using Grid's tools as part of our end-to-end solution that is streaming DVD quality video over the web, to customers all over the world.

Second part coming soon.

Innovation #1, I liked in Japan

2006-12-07T21:12:00.000-08:00

A much needed accessory for the toilet.

Challenges of High Quality Video Delivery

2006-12-06T22:24:00.000-08:00

Since Google acquired YouTube, online video sharing and delivery segment has been hot topic of discussion. A good overview of online video viewing is presented by Scott Kirsner in As online viewing booms, the amateurs give way to big media. As more and more big media content coming online, a new technology challenge is emerging for content distributors.

How to deliver a high quality full-length video instantly to multiple viewers on their big screen simultaneously and securely?

Sneakernet

The file size for 90 minute full length HD quality video can range from 2 to 6 GB. And it may be easier, convenient and hassle-free for viewer to get it delivered via next day Sneakernet than to wait 3 to 18 hours in downloading it online. There is nothing instant about downloading full-length high quality video.

Streaming

The next best online alternative is to stream video to viewers in real-time. For smooth playback, video streaming needs to have some buffering and a 12 second buffering of HD quality video requires a 5 to 15 MB of video that need to be always available to user. And, if streaming video is originating from a centralized distribution infrastructure, the number of viewers are first limited by the processing capacity of central system and then available bandwidth.

Content Delivery Network (CDN)

To address the central system processing capacity limitation, the option is to have multiple delivery nodes. And, to address bandwidth limitation, these content delivery nodes should be located as close to the viewer as possible. This is the model used by CDN providers who share the processing capacity of these nodes and available bandwidth among multiple content distributors to maximize utilization of their edge nodes.

Content delivery nodes work great as most viewers reside on the last mile that extends beyond the Internet spiderweb and as long as nodes are not overloaded by too many viewers trying to download and watch the latest video simultaneously.

Peer-to-peer (P2P) Networking

P2P networking is one scalable alternative proposed to deliver high quality video and bring the viewers in to the Internet spiderweb. Among it's other vices, from delivery perspective, P2P is infested with freeriders. You know the ones who want to download content but don't want to allow their system to be used to deliver content to others. And, the whole P2P premise fails if there are not enough P2P nodes participating in content distribution.

GridNetworks – New Kid on the Block

Today, I had the pleasure of meeting and talking with Newell Edmond, co-founder and the technical brain behind a local startup in online delivery of high quality video, GridNetworks. Newell and his team worked hard for last two years to develop a hybrid solution combining the best of Content Delivery Network (CDN) and Peer-to-Peer Networking (P2P) for delivery of high quality video.

More about it next time.

P.S. This is my first effort to reach out and highlight early stage startups in Seattle area working on innovative infrastructure solutions. If you are one of them, get in touch.

Rent - A - Blogger

2006-12-04T17:56:00.000-08:00

Last week, NetGear announced contest for a guest blogger who want to blog for them at CES. Contest detail links at Jeremy Toeman blog.

After reviewing the official rules, my impression is that a journalist/media writer who is accustomed to deadlines for articles or someone who blogs often during the day will fit their requirements better. Most probably, NetGear will provide product briefings as their expectation is to blog how they fit in to the other stuff showcased at CES.

Why go outside the company to look for a guest blogger? Other than the marketing buzz about contest announcement, what value does he or she provide? You are unlikely to bag a high profile blogger, like Robert Scoble, with such gimmicks.

Guest blogger gigs can be a good avenue for niche bloggers like me who like to attend few relevant industry events but typically don't get sponsorship by their own organization. I was thinking of attending USENIX Conference of File and Storage Technologies (FAST) San Jose, CA in February. May be I should rent myself out as guest blogger for the event? Anyone interested?

You got any thoughts about guest-blogging and sponsorship, chime in.

Everybody who saw this picture, asked me what it is? I captured this image near Poplar Avenue at Hokkaido University. What is it?

Comment on A-List Blog may result in high traffic

2006-11-30T20:24:00.000-08:00

Like most other bloggers, I also keep track of the traffic on my blog. I credit Jeremiah, ex-Hitachi web honcho, for suggesting to keep traffic stats using Google Analytics. I noticed in stats for my blog, that almost half of traffic on Oct 24th came from one source, Scobliezer, an A-List blog by Robert Scoble. No, I am not one of those lucky ones that were mentioned by Scoble.

Most of the traffic originated from his blog entry Internet video business challenges. One of the point, he made in his post was about the high distribution cost with delivering high quality video. He mentioned that he is going to collect $10 in advertising to pay $28 in bandwidth cost.

In my opinion, development cost is one time cost for creating one high quality video but bandwidth cost is proportional to number of times that video is downloaded from fully-owned single distribution point. And there is no easy practical way to reduce the delivery cost without deploying a multi-owned multi-point distribution network. So, I wrote this in comment section (#32) of his blog entry.

Robert,

Mark Cuban is right on target. I will take his idea little further.

Why not allow users who download your video to share and transfer it to other users? This way, you will not need to purchase more bandwidth to txfr video to more users.

Figure out a way to manage digital rights so that even though video can be transferred from user to user but it can’t be played without a digital license. And the license is only distributed from your site.

Combine the digital license with new relevant ads that get embedded with video.
Google (someone else info, my ad) + peer-to-peer bulk data transfer (Napster, only legal) = Ad revenue + Infrastructure cost saving.

Anil

And, this is what Scoble wrote in his comment (#34) as a response to point I raised:

Anil: using BitTorrent or other P2P distribution schemes (RedSwoosh) is very interesting to me. I’m definitely looking into those.

Time to time, though not often, I do write comments on other blogs including A-list blogs typically resulting in little or no impact on traffic to my blog.

So, what was different this time that resulted in higher traffic?

I can only guess two possibilities that caused people to click on my name and check out who I am, the idea mentioned in my comment or validation by Scoble to consider it as an interesting alternative.

Now it raises another question. When making comments on blogs, what is more important - quality, quantity, or frequency? I prefer quality over the other two.

And, hopefully this post will give a pause to someone to rethink his strategy of making irrelevant comments on blogs to promote his storage job blog.

A toilet/restroom in Hokkaido constructed using marble. It also comes with a Yamaha piano, playing on its own, to entertain you while you go about your *smelly* business!

Impressions of Storage and Service in Japan

2006-11-29T19:44:00.000-08:00

Most of the time, I was a tourist in Japan. But once in a while I also had serious discussion with whoever spoke better English than my Japanese or put up with an interpreter who only translated selectively! Some discussions were about population trends, storage market and opportunities and service nuances. Whether it was Sapporo, Kyoto or Tokyo, I was surprised and impressed that most Japanese, I talked to, considered issues with worldwide perspective instead of what is good for Japan unlike some other places.

Storage professionals seem to be as much in demand in Japan and APAC as in North America. I wish I had some Japanese language skills, I would have easily skipped my return flight. ;-)

I also heard similar sentiments about Hitachi storage as I hear in North America about its mindshare and footprint. Main difference being that most comments in Japan were directed not at HDS but at Hitachi's disk drive unit. HGST may be the better known entity in Japan than HDS. Where is storage on its list of priorities at Hitachi? Chime in if you desire. And, what's up with everyone wanting to discuss HDS with me?

I was particularly impressed with the level of service extended to customers in Japan. I don't think I have experienced such service dedication anywhere else. One individual from auto industry put it nicely "Here you get better service because it is ingrained culturally. In US, you get good service because people get paid for it." Just makes me wonder why HDS is generally thought of as "great hardware, ok software, poor service" company.

Promotional display for ASUS-Lamborghini Golden Edition laptop in Akihabara area of Tokyo. If you are in the area, try visiting Yodobashi Camera, a huge discount electronic store. I haven't seen anything like it on this side of Pacific.

Thanksgiving in Kyoto

2006-11-27T19:57:00.000-08:00

I hope my readers had wonderful Thanksgiving holiday. I spent my thanksgiving in Kyoto enjoying fall colors at Kiyomizudera temple and eating Puffer fish, a Japanese delicacy, for dinner.

I enjoyed my visit to wonderful Japan. It was also a nice break from laptop, cell phone, Internet, email and blog.

Now it is time to get back to storage.

Be Generous

2006-11-12T22:29:00.000-08:00

Busy preparing for the trip to Japan. Already in holiday spirit. Need to catchup on storage stuff.

Medina House of an Hungarian guy. He reportedly received one percent stake to leave IBM and work for a startup, now we know as Microsoft.

Got photo blog inspiration from Andy Gray's Japan Photo Blog.

What Do You Care What Other People Think?

2006-11-05T22:31:00.000-08:00

Last few weeks, I spent some time exploring life and work of Richard Feynman (1918 - 1988), a winner of Nobel Prize in Physics, inventor of Feynman diagram for quantum field theory calculations, a member of the investigating panel on Space Shuttle Challenger disaster, and a professor at Caltech.

Read book Surely You're Joking, Mr. Feynman! Adventures of a Curious Character Richard P. Feynman

Listened to an audio book What Do You Care What Other People Think? Further Adventures of a Curious Character Richard P. Feynman

Watched a PBS NOVA documentary DVD The best mind since Einstein profiling Richard Feynman.

Watched this 50 minutes Google Video The pleasure of finding things out - RP Feynman

Why did I want to know more about Richard Feynman? He was brought to my attention by my boss. For a while, he has been commenting that I remind him of his Physics professor at Caltech. Now, I am sure not for my beautiful mind but for my eccentricities like speaking my mind and questioning status quo.

In the end, learning about Richard Feynman and his work gave me a new perspective on feedback received from others irrespective of whether it is about my blog or the ways I approach and discuss issues.

Thank you all for your opinions and feedbacks and keep them coming! And, no I am not growing up and I have no plans to quit blogging.

Data De-duplication crossing the Chasm

2006-11-02T21:12:00.000-08:00

For last couple of days, storage blogosphere and media has been buzzing about EMC acquisition of Avamar, a data de-duplication vendor (See related entries below for the extent of coverage). It wasn't a best kept secret as rumors were circulating for a while that Avamar is being courted by EMC.

Recently, I have been studying Avamar data de-duplication technology after writing several times about data de-duplication and Data Domain. Few weeks ago, they were kind enough to send me promptly marketing and technical material. Don't know if that had to do with me being storage blogger or their openness to anyone interested.

Congratulations to Avamar for a very attractive exit. In exchange, EMC is getting a strong technology platform in real-time data de-duplication and handling both data-at-rest and data-in-transit. And it has potential to take EMC's backup, archive, replication and, contrary to popular thinking, array products to the next level.

Acquisition of Avamar by EMC, Rocksoft by ADIC, rumored IPO plans of Data Domain, and total segment revenue potentially reaching $100+ million, it appears that data de-duplication finally has crossed the chasm.

Related Entries:

Jeff Boles, Tony Asaro, Storagezilla, Storagezilla again, Mark Lewis, Byte & Switch, Dave Raffo, Beth Pariseau, Sharon Fisher, Raj Bala, Joseph Kovar, Tim Stammers

Preserving Wynton Marsalis in Digital World

2006-10-25T20:22:00.000-07:00

Last weekend, I saw Wynton Marsalis perform at Paramount Theatre. During the performance, I started to think about how legacy of Wynton Marsalis would be preserved in digital world.

Long time ago, the primary way to preserve music was to write on music sheets. Overtime, some of these sheets were physically destroyed or ink faded. One way or the other, some of the music was lost. Also, different people interpreted the written music differently (A debatable point as I am no expert on this topic ... treat it as view from the outside) as there was no audio or video archive of the original performance. The main challenge in that era was physical preservation of music sheets.

Things in music world have changed dramatically since the introduction of audio and video technology. Now, music is primarily preserved using audio, video or both. This shift has allowed more and exact information about a particular music to be preserved like the interpretation of the music by the original artist and his or her original performance. With digitization, even the making of a particular piece of music can be archived for future generation. This change also has brought new challenges in preservation of music.

Will we still have access to the music of original artists, hundred or even twenty years from now? The answer is no longer as simple as physical preservation of music sheets. The digitization of music along with its benefits has brought new challenges in long term preservation of music.

Survivability of the audio and video file formats.

Capability of the application to play original audio and video file formats.

Availability of original digital rights management process if DRM was applied to the file.

Availability of original encryption key management system if file was encrypted.

Capability of the operating system to be able to host application and support media and system.

Physical survivability of the media on which audio and video are recorded.

Physical survivability of the system capable of using the original media.

Come to think of, the long term preservation challenges are not exclusive to music, they also apply to all other information that need to be preserved for the future generations. And, these issues become lot more complex to address once the file, application, keys (DRM, encryption), OS, media, system, and storage are physically separated.