Zerowait Corporation

I was not surprised to read this article on Byte and Switch today. I have long thought that performance and Benchmark numbers were suspect in the storage and networking world.

Highlights of the article:
* Driven by a general sense that benchmarking practices in the areas of file and storage systems are lacking, we conducted an extensive survey of the benchmarks that were published in relevant conference papers in recent years. We decided to evaluate the evaluators, if you will. Our May 2008 ACM Transactions on Storage article, entitled “A Nine Year Study of File System and Storage Benchmarking'”, surveyed 415 file system and storage benchmarks from 106 papers that were published in four highly-regarded conferences (SOSP, OSDI, USENIX, and FAST) between 1999 and 2007.

Our suspicions were confirmed. We found that most popular benchmarks are flawed, and many research papers used poor benchmarking practices and did not provide a clear indication of the system’s true performance. We evaluated benchmarks qualitatively as well as quantitatively: we conducted a set of experiments to show how some widely used benchmarks can conceal or overemphasize overheads. Finally, we provided a set of guidelines that we hope will improve future performance evaluations. An updated version of the guidelines is available.

We believe that the current state of performance evaluations has much room for improvement. This belief is supported by the evidence presented in our survey. Computer Science is still a relatively young field, and the experimental evaluations need to move further in the direction of precise science. One part of the solution is that standards clearly need to be raised and defined. This will have to be done both by reviewers putting more emphasis on a system’s evaluation, and by researchers conducting experiments. Another part of the solution is that this information needs to be better disseminated to all. We hope that this article, as well as our continuing work, will help researchers and others to understand the problems that exist with file and storage system benchmarking. The final aspect of the solution to this problem is creating standardized benchmarks, or benchmarking suites, based on open discussion among file system and storage researchers.

Our article focused on benchmark results that are published in venues such as conferences and journals. Another aspect is standardized industrial benchmarks. Here, how the benchmark is run or chosen, or how the results are presented is of little interest, as these are all standardized. An interesting question, though, is how effective these benchmarks are, and how the standards shape the products that are being sold today (for better or worse).

The goal of this project is to raise awareness of issues relating to proper benchmarking practices of file and storage systems. We hope that with greater awareness, standards will be raised, and more rigorous and scientific evaluations will be performed and published. Since this article was published in May 2008, we held a workshop on storage benchmarking at UCSC, and we presented a BoF session at the 2009 7th USENIX Conference on File and Storage Technologies (FAST). We have also set up a mailing list for information on future events, as well as discussions. More information can be found on our Website, http://fsbench.filesystems.org/.

Over the holiday weekend I got together with a bunch of my friends who own businesses or are executives in companies. We had a chance to talk about what they see in their businesses and there was not a lot of optimism for economic growth in the USA. Many of the companies they represent have stabilized their businesses, but they are not expecting domestic growth this year. The consensus was that a lack of Capital Expenditures is forcing companies to spend more on maintenance instead of upgrades and replacements, but even operating budgets have been curtailed.

My informal research coincides with a report I saw by Reuters today. Some points in the report are below.

* The services sector represents about 80 percent of U.S. economic activity, including businesses such as banks, airlines, hotels and restaurants.

* Both manufacturing and service sector reports show signs that the 18-month-old U.S. recession, the most protracted in decades, may soon end.

* The Conference Board, a private research organization, said its Employment Trends Index slipped to 88.4 from a downwardly revised 89.1 in May. It was originally reported at 89.9.

* “Compared with the beginning of the year, the decline in the Employment Trends Index has significantly moderated,” said Gad Levanon, senior economist at the Conference Board. “We therefore expect job growth to resume around the end of the year. “However,” he added, “over the last month, leading indicators of employment were mostly disappointing, suggesting the Employment Trends Index is still seeking a bottom.”

Green Shoots or Dead Roots – I think it depends on what sector of the Economy and what business you are in.

According to an article by Chris Mellor :

The Enterprise Strategy Group is forcasting a sixfold growth in file archive capacity, from a little over 10,000PB in 2008 to 62,000PB in 2012

This seems to one of the few bright spots I have seen in the last few months when it comes to economic growth . I see a few interesting points that this growth will have on ever tightening IT budgets.

1) If budgets remain tight, customers will have to seek cost savings on legacy equipment to purchase new hardware to handle the growth. Many OEM’s raise support prices on legacy equipment as an incentive to current customers to purchase new equipment. Since budgets are tight customers will be seeking more third party support to reduce their operating expenses on their legacy equipment. That should be good for Zerowait.

2) Customers will be looking for better ways to manage their RAID arrays to gain more efficiencies out of current hardware. Whether they choose compression techniques or de-duplication technologies, Storage Resource Management is still going to be a headache.

3) My recent overseas trip clearly demonstrated that customers around the world are seeking ways to squeeze more out of their storage budgets. Large international companies and smaller regional enterprises are looking at which state or country is offering the lowest cost data center.
Clearly storing and serving more data is going to take more power, and therefore there are advantages for storing data in locations that offer cheap, reliable and abundant electrical power and connectivity.

Most of the companies we deal with have three tiers of data – Primary, Secondary and Tertiary data sets. Primary data needs to be ‘connectivity near’ the users but as data gets less critical it can be ‘connectivity farther’ away.

In the future we may see more data centers in low cost locations being filled with high availability legacy equipment that is cheaper to maintain than new equipment for the secondary and tertiary data sets.

Economics will force folks to ask “why are we serving data from the most expensive locations?”

If headlines like the one below continue companies are going to have to cut costs to survive:

Unemployment flirts with an ugly truth–the possibility of hitting a new post-Depression high.

Over the last couple of weeks I was visiting customers internationally and I was surprised at how many customers are trying out www.isilon.com . The reason they are trying Isilon is price, or at least that is what I was told. According to the customer comments I heard, NetApp seems to be pricing itself out of these high performance storage markets. Whether this is motivated by customer budget decisions or a NetApp international pricing policy I can’t tell. But for whatever reason, Isilon has now got a foot in the door of the data centers I was visiting.

I was surprised not to see EMC and IBM making more headway into NetApp’s international customer base, since Isilon is. In times of economic turmoil I would expect customers to go to the safety of a major brand, but if price is the major motivator of these decisions, Isilon may be buying customer loyalty. The installations are too recent to have meaningful long term reliability results.

I don’t run into Isilon very often in the US and Canada, so it might be that Isilon is targeting NetApp’s international customers. Or it could just be coincidence, and there is no correlation at all.

Sometimes technology markets work in mysterious ways, and the storage market is currently reinforcing its reputation.

Over the last couple of weeks I have been traveling to visit customers in many locations. Can you guess where I was last week from the picture?

Our domestic and international customers are looking at how to squeeze more out of their IT budgets and have been asking us to provide support for other parts of their networks. We are reviewing their requests and looking at whether we can provide high availability support for more of our customers’ network components.

Providing outstanding High Availability support is challenging, although we must be doing a good job as our customers are requesting us to take over support for additional hardware within their networks.

Often the trade press is filled with regurgitated press releases which don’t give a very good analysis of the companies and strategies that are winning customers and markets. But I just read a very good article that I would suggest you read.

NetApp has been losing both mind and market share to the more aggressive EMC, which has partnered with Cisco. With Oracle having acquired Sun (and with it, Sun’s storage business) and HP expected to start building its own storage systems, there are few takers for NetApp. As a result, the company’s only option was to start buying up smaller but fast-growing players such as Data Domain. For the longest time the knock on NetApp was that it wasn’t able to buy and digest companies, as best evidenced by its struggles with its Decru and Spinnaker acquisitions. Now it’s being straight-up outbid by EMC, which is offering cash. EMC, sensing blood, wants to run NetApp out of town.

Our business is growing in the tough economy because individuals and the market are recognizing that there is value in maintaining their reliable legacy equipment. The battle for Data Domain may be a harbinger of things to come. But my crystal ball on this subject remains cloudy.

But as budgets remain tight, I expect more companies to ask for lower cost service and support for their enterprise storage systems.

According to many analysts the enterprise storage market has been shrinking and NetApp and EMC are slugging it out for market share.

The NAS (network attached storage) market declined 6.7 percent year over year, led by EMC with 39.0 percent revenue share and followed by NetApp with 28.7 percent share.

Coincidentally, we are seeing growth in our service and support business this year, as companies look to save on maintenance and extend the lifespan of their legacy equipment. Extending the time between upgrades is one way to get more value out of your storage infrastructure, another way is to manage your storage resources better.

Creating a plan by which you can consolidate your existing storage, and reduce service and support expenses can stretch your budget dollars. That is why more companies are coming to Zerowait and asking us to provide them a free storage analysis and quote on their current storage infrastructure support. Superseded High availability storage equipment can still provide many years of service to most organizations for a fraction of the price of an upgrade.

Data Domain may be the most popular company in the Storage niche in the last few days. In hard economic times it pays for customers to want to save money by increasing their storage efficiency and de-duplication makes a lot of sense if a customer can see real cost savings.

As a larger company with more large storage customers, EMC has a bigger market than NetApp does within its existing customer base. And they may be able to use DD to help their customers save money by reducing their storage requirements. Also, since many DD customers are also NetApp customers it gets EMC into the door of existing NetApp accounts.

EMC seems to have a better history of integrating companies than NetApp does, and so from a customer and employee point of view folks may want to see EMC win this battle.

During a late Monday afternoon press conference, EMC CEO Joe Tucci promised that his company would run Data Domain as a product division within EMC, and would increase its R&D investment in Data Domain technology. “We look forward to giving the Data Domain employees a warm and heartfelt welcome into the EMC family. I would like to remind everyone EMC has a good track record of integrating companies, retaining and growing key talent and achieving results,” Tucci said. The CEO didn’t mention the less-than-stellar integration of VMware Inc. technology into EMC, though to be fair, that acquisition has been a financial success so far.

On the other hand, NetApp does not have a good track record with at least some of its acquisitions, notably the Spinnaker Corp. deal in 2003 and the acquisition of Decru Inc. in 2006. Both companies went through a very rocky integration road with NetApp. “I do happen to believe that NetApp does acquisitions miserably,” said Arun Taneja, founder and president of the research firm, Taneja Group Inc.

De-duplication is another patch on the storage resource management problem. Storage is growing and as it grows it gets more complex to manage, de-duplication shrinks the footprint but adds another layer of management.

De-duplication is not the answer to the storage resource management problem.

Over the last week many people have asked my opinion of this acquisition. I was not surprised that NetApp purchased Data Domain, I know quite a few former NetApp folks who work there, and it will be interesting to see how they are re integrated into the NetApp fold. I think Ken Hibbard is perhaps the most senior former NetApp employee to be rejoining the company .

From Hoovers. com Mr Hibbard comes to Data Domain from Network Appliance, where he was vice president of product engineering. He has joined Network Appliance in 1998 and was named vice president of software development in February of 2000. In January of 2001, he became vice president of product development and East Coast operations.

It looks as if the Data Domain folks understand the NetApp product development cycle and therefore may be able to integrate their products into the NetApp Line card a lot quicker than the Spinnaker products have been integrated. I certainly hope that the integration of Data Domain’s customers and employees works out better for them than the Topio acquisition did for their customers and employees.

NetApp is closing down its Snap Mirror for Open Systems product, which came as the main part of its $160 million Topio acquisition.

Topio was bought in November, 2006, and its main product was an any-source-to-any-destination replication product called the Topio Data Protection Suite.

According to an article in CXO

... Data Domain lacks the customer base of large vendors and has been facing market challenges competing against Quantum.

In today’s environment customers are looking for the best value they can get for their storage dollar. The question will be if NetApp can sell Data Domain products at an aggressive enough price point, after paying such a premium for the company?

Over the next few months it will be interesting to see if NetApp can integrate Data Domain offerings into their product line and execute a sales strategy that works better than their recently ended Storevault strategy did.

6:31 PM EST Tue. Feb. 03, 2009 NetApp has quietly decided to kill its line of storage appliances for SMBs, a move which caused mixed feelings in the storage vendor’s solution provider community. NetApp told some users and partners that it has decided to end the availability of its S550 platform, which was the last product in its StoreVault line of appliances. NetApp unveiled its StoreVault line in May of 2006 as a way of entering the SMB market.

The EMC World show was great. From a trade show marketing point of view it was probably one of the best shows we have ever attended. The attendees understood the storage industry and were looking for answers to their storage and budget issues. We saw many of our old friends within the industry and a few of our customers at the show.

The EMC folks put together a well organized and smoothly run trade show.