Aperi: All Talk, No Walk

I was just reading comments written by Tarry in response to my posting Which industry segment most likely to use Virtualization? He mentioned about APERI project by IBM et al. Jon William Toigo also has been blogging and writing columns about Aperi.

My (mis?) understanding is that Aperi is suppose to be a "open source" initiative. When was the last time a "proprietary source" company started a successful "open source" project?

It has been few months since Aperi announcement was first made. So, I figured by now there must be some real stuff out there about this "open source" initiative. And what did I find? Nothing, nada! Only press releases and media articles. Even website of IBM, supposedly the leader of this gang, has no substantial information on Aperi. Toigo's blog gives the impression that Aperi initiative has been dumped now in the lap of SNIA.

I am impressed with the progress of the Aperi project and the participants' commitment. Can someone care to explain, why should users buy into Ideas running on fumes?

Coolest Product, Episode 2: Technology

Today, while reviewing the new blog posts, I came across Heidi Biggar entry To de-dupe or not to de-dupe about data de-duplication / capacity optimization / commonality factoring / single-instancing. In my last entry The Coolest Product from a Storage Startup, I mentioned coming across an exciting product and this product happen to fall in to the same category as mentioned by Heidi.

I agree with her observation that capacity optimization will be one of this year's most talked about new technology. Unlike storage virtualization hype of today, capacity optimization is delivering measurable and significant value to customer. But, it may take few years before the full potential of this technology is realized.

I believe the impact of capacity optimization goes beyond just data protection to primary storage. Just consider the possibility of able to store 50TB data on a 3TB storage array. The technology is quite powerful specially when it is positioned as data virtualization across an enterprise - removing the repeating segments everywhere in the enterprise with pointers to the centrally stored single instance of that segment.

A good "geek" read on the inner working of this technology is the US patent 6,928,526 Efficient Data Storage System granted to Dr. Kai Li, co-founder of DataDomain, and others. In my opinion, they shortchanged themselves with restricting their patient to data protection only or may be that was the only way to get the patent as I noticed patent examiner added several other patents as reference. I haven’t had chance to read them yet.

I fully agree with statement by Heidi about capacity optimization:

I'm pretty certain you'll be amazed at the results!

The Coolest Product from a Storage Startup

Suspense music starts ...

Couple of days ago, I came across an interesting product being evaluated in a data center. The data center personnel made very encouraging comments about this product. Since then, I have been researching the product with intention to write a blog entry.

Tonight, I noticed Dave Hitz comment The Coolest Storage Vendor. NetApp may be the coolest storage vendor according to Tony Asaro (see Storage Jumboree) but this product may be the coolest product, I came across recently, from a storage startup ... initial impression.

Opinion subject to change as I haven't yet finished reviewing all available information. ;-)

<< More to come >>

... suspense music fading away.

Which industry segment most likely to use Virtualization?

Last week, I visited a large data center that houses hardware for various customers. I came across a very interesting cage while walking along a row of cages. This cage contains over thirty racks full of servers, networking and storage hardware.

The best, I could observe, it has storage from NetApp, EMC, HDS, HP, Dell, and Adaptec; networking from Cisco, Extreme, F5, Netgear, and Brocade; server equipment from Dell, Sun, HP, SuperMicro and other whitebox vendors. They are running Solaris, HP-UX and Linux as guessed from rack labels.

The two things stood out in my mind:

• Absence of IBM equipment in the mix.


• The hodgepodge of vendors whose equipment are installed in this cage.

It is rare for me to see this much heterogeneity in an environment. I typically recommend no more than two or three vendors for competing products form operation management perspective. I also suggest not standardizing on one vendor platform to avoid losing the future leverage. Actually, I learnt this over 10 years ago working for a very large company - two vendor strategy … funny how the lessons learnt in early stages of career are so difficult to forget.

Is this organization a prime target for virtualization? I am pretty sure in near future; either this company will be a good candidate for Virtualization or retiring few redundant vendors to simplify their environment.

We all hear vendor marketing about how virtualization can simplify storage management. But the truth is unless you inherit a messed up environment, the virtualization may introduce more operational complexity instead of simplifying management. And even in this scenario, the cost-benefit analysis must be performed to evaluate the benefit of virtualization versus scaling back on number of vendors in the environment.

Looking at multiplicity of vendors in this environment got me thinking about which industry segments are most likely to use virtualization. The organization, who owns equipment in this cage, seems to have an ad-hoc hardware acquisition strategy. Considering they are in a managed data center, I will guess that it is a web based business or a managed application service provider.

1. Organization that has gone through several M&A in last few years, especially merger of equals.


2. Geographically dispersed organization with de-centralized management.


3. Organization whose management seem to be always in structural flux.


4. Technology industry segment.


5. Organization that grow too quickly too fast, startups like Six Apart. Actually Om Malik of Business 2.0 had an interesting post last month on his blog about the infrastructure needs of rapidly growing Web 2.0 startups. See Even in Web 2.0 Scale & Size Matter.

What are some other qualifying criteria for the organizations that may be potential users of virtualization?

I am looking forward to heading back to Toronto later this week.

A Blogger Reports Record Q4 Revenues $168.64

It is just my luck, trying to get away from snow in Toronto, landed up in the middle of twenty-four consecutive rainy days in Bellevue, Washington. Today, finally I got the chance to catch up on reading some of my favorite blogs.

Yesterday, Bill Burnham, a venture capitalist with infrastructure software focus, wrote a hilarious post Burnham’s Beat Reports Record Q4 Revenues parodying the typical press releases by companies announcing their quarterly earnings. Take a break from boring storage business (unless you lost your tapes) and have some fun reading his post.

This quarter’s results continue to demonstrate that blogging is a complete waste time.

Fortunately, my blog revenues, expenses and income are exactly same $0.00. No revenues as I don’t have any advertisement or affiliate programs. No expenses as I am cheap and use free Blogger to host my blog. And, sweat equity doesn’t count for even a penny. No performance metrics to report as I don’t track and maintain any metrics. I guess I am too busy with other things and too lazy to do anything about it.