Zerowait Corporation

Yesterday, we received a call from a customer complaining about how unreliable the Maxtor drives were in his R150. We had a long discussion about the reliability problems of the drives and he could not understand why NetApp uses unreliable drives. He had been reading my blog about the BCS and ZCS issues and was also steamed that NetApp is selling drives that are less reliable and have less error checking than the drives they used to sell.

I explained that he was not alone, and we agreed that NetApp is much more focused on profitability then reliability. Anyone can read NetApp’s founder’s Blog and see that he writes about the size of the company, moving engineering to India, and avoids issues related to reliability.

But what do I know? NetApp is growing, so perhaps customers don’t care about reliability anymore.

Do Seagate drives sold by NetApp have some mystical powers? Not according to Hitachi!

I have been confused for years about the benefits of 512 (ZCS)Sector drives as compared to 520 (BCS) sector drives. And as one of our clients pointed out the R200 uses the older inferior 512 sector (ZCS), according to NetApp ZCS technology is inferior. So why would a newer unit, for example the R200, use superseded technology? I was confused and could not find an answer to this question from anyone at NetApp or from anywhere on their website. So I wrote a note to the guys at HDS. Since HDS works with NetApp on their Gfilers, I figured I could get a clear answer from them.

According to the HDS engineer
“There is no measurable difference in reliability between 512 byte and 520 byte sectors.
SCSI and FC drives can be low-level formatted with different sector sizes (within a small range with 512 at the low end).
The reason for this is that the payload of 512 bytes is sometimes augmented, as it is in Hitachi’s case on the RAID and DF, by additional check byes appended to the end of the sector. I think Hitachi calls these CRC check bytes, but they also identify which sector it is, (perhaps by an “ID-less” technique where when the CRC bytes for the 512 byte sector are computed, a virtual ID field is included in the check bytes. That way, that sector will only read correctly when it is read from the right place.
So the bottom line is that this is a case of a subsystem vendor applied redundancy code applied on top of what the disk drive already supplies.”

It leaves me with a feeling that all this drive jargon that NetApp has been marketing to us for several years is simply marketing spin. And another way to make their systems more proprietary.

Comments – from Jon Toigo on this post

Hey Mike, the really interesting thing to me is that NetApp contacted me through two different routes to describe how popular their solutions are for data protection. One fellow claimed that their gear was more popular than anyone else’s for disaster recovery. When I invited them to make their case as a sponsor of the Disaster Recovery and Data Protection Summit, they demurred. Guess they were afraid that you would be there…

Jon Toigo
Toigo Partners International
Managing Principal
Data Management Institute

Founder
1538 Patricia Avenue
Dunedin, FL 34698 USA
O: 727-736-5367
F: 727-736-8353
C: 727-504-9311jtoigo@toigopartners.com

Does NetApp charge so much for its software because customers have incomplete information to make an informed decision?

According to an article in the Economist April 15, 2006 edition on page 78 – Physicians and Lawyers wives have much fewer surgeries than the general population. The article says that this is because Surgeons have better information than the general population, and that Surgeons fear operating on Lawyers wives. This same article points out that Real Estate agents on commission sell houses cheaper than they would their own house.

If customers had complete information about NetApp’s hardware and software policies inevitably their cost of storage would fall. And while there is competition for NetApp in the general storage market. Once a customer buys into the NetApp value proposition, Zerowait is the only viable option for transferable licensed filers , and service and support for NetApp’s equipment.

The best time to purchase from NetApp is at the end of the quarter or the end of their fiscal year. This is because the sales department wants to make as much money as possible so they will cut their prices to sell more equipment. Additionally, the NetApp salesperson is into his Commission Accelerator period at the end of the year, So his commission rate is higher. Just like a real estate agent NetApp’s sales folks will cut price to sell more product, and raise his income.

As NetApp’s end of year is fast approaching, the best way to get price concessions from NetApp will be to show them competitive quotes. Either from Zerowait or from one of the other storage vendors. And remember to insist on getting your transferable software licenses from NetApp.

History shows that where ethics and economics come in conflict, victory is always with economics. B.R. Ambedkar

Why does NetApp tax their users more to manage 24TB of data on an R200 than 12TB of data?
In short NetApp charges more because they think they can, but in the long run they will lose customers to lower cost alternatives. It is the way of the world.

One of our customers asked us this the other day, and I had to tell him that I could not figure out a reason NetApp would charge more to manage more data. It is the same software, and it has the same amount of programming. So why does it cost more?

Perhaps NetApp views software the way the IRS views income? The more you have the higher the tax on it. It seems ironic that NetApp would put a progressive software tax on storage, since the one thing all enterprises want to do is lower their marginal tax costs.

How can a NetApp user lower their marginal storage tax burden? There are a few of ways:
1) Purchase transferable licensed systems and then max out the storage on the systems.
2) Use storage systems that don’t tax storage as much as NetApp does.
3) Get a quote from companies like Zerowait and show it to NetApp, that usually has a downward effect on NetApp’s priocing.

Over time NetApp’s high marginal tax rate on storage will cause customers to look to other sources of storage which are cheaper to maintain. Think of it as the gloablaization of storage. Just as NetApp is using India for engineering, customers will sooner or later embrace lower cost alternatives to NetApp.

Why does an R200 Nearstore use 512 Sector drives? What is going on? Maybe the reason that Dave Hitz is advocating Dual Parity now is because NetApp is using less reliable drives and drive technology?
A few years NetApp swtiched to 520 sector drives because they were more resilient . But on their Back up and Archiving units they currently use 512 Sector drives. Are the Maxtor ATA drives less error prone then the Seagate FC drives?

NetApp Disk Formatting: ZCS and BCS
Network Appliance filers in the early days came with with “normal” 512 sector disk drives. But the drive firmware’s built-in check summing occasionally caused parity inconsistency, or data corruption.

So NetApp created their own method of check summing and introduced it in Data ONTAP 6.0. With this new method, every 64th strip on disk was a check sum of the previous 63. Unfortunately, performance suffered terribly, some say up to 30% overhead, until the engineers remembered the mainframe technology of old, where there were 520 sector disk drives with the check summing built right in to each sector. No more 64th strip and the performance was vastly improved.

As a result, the old, “normal” 512 drives became ZCS or Zone Check Sum–named after how the program slips the drive into zones of data stripes and check sum stripes. The new drives with the 520 byte sector became BCS or Block Check Sum drives because the check summing is “built-in”.

MAY 31 & JUNE 1 – Join us in Tampa!

Jon Toigo’s group is putting on a Disaster Recovery Conference and we will be there! Although we like to advocate Disaster Prevention techniques, rather then waiting for a problem to occur.

It looks like it is going to be quite an event and the interest that we are getting from our customers is quite good. I think a good cross section of our customers will be there. If you want to attend you can go to the website that I have linked to at the top of this entry.

I hope to see you there!

It has been a very busy couple of days. We have started to get in NetApp FAS960’s and a lot of people want to upgrade from their 800 series units to the FAS960’s, and have been waiting for our stock levels to increase to satisfy these customers. Additionally, now we are geting calls from NetApp customers who are looking at new systems. Typically, a NetApp customer gets a quote from NetApp for a FAS3020 looks at the price – faints – recovers – and gives us a call – we tell them to go to www.spec.org and look at the performance of the FAS3020 or FAS3050 and compare them to the specs on the FAS960. Many are left wondering why they should buy a 3000 series unit from NetApp. We have been trying to understand the reason to upgrade from a 940 to a 3020 for many months. But, because of recent articles in the trade press end user customers are also scratching their heads.

Additionally, our customers are embracing our ZHA exception reporter. During this morning’s staff meeting I was told that yesterday my staff received two calls from customers who are using it to improve their filers performance. If you are using NetApp filers, you should try our exception reporter, it is really a useful tool.

Morphic Resonance

Last night I phoned Jon Toigo because of the search storage article and we chatted about it at length. This morning I read his blog. It seems that the press and the competition is now starting to ask the same questions that Jon, Zerowait, and our customers have been asking for a couple of years now.

When is NetApp going to provide reliable, repeatable and verifiable performance numbers? Competition in storage is strong and growing, and I predict that NetApp’s well funded competitors will soon start to push for Verification of NetApp’s claims.

It is not only Zerowait and our customers who are looking for real numbers, but also the trade press.

NetApp – we used to be a contender

It is eerie to watch the decline of NetApp. Here is company that spends hundreds a millions of dollars to purchase Spinnaker, Alacritus, Decru and yet does not spend money on listening to its customers. It is like watching GM’s decline in many ways. GM thought that by purchasing Hummer, part of Fiat, and Saab they could stop the decline of their company. But the problem is not the brands it is that the company does not listen to its customers. NetApp likens itself to a New Economy company, but it is really the same old story. Listen to your customers, they will tell you what they want and what they are willing to pay for it.

So what will happen over the next few years? Competition will force NetApp to lower its prices, I predict that because they have a very high overhead they will slowly cut into their currently very high gross margins. As their current customers smell that their salesman are getting more desperate the customers will demand higher discounts. This will cause NetApp to reduce its workforce and offload unprofitable divisions and projects. Will NetApp’s shareholders ever see a return for their investment in the Spinnaker purchase? I strongly doubt it.

It will take a couple of years, but competition will reduce NetApp to a memory, just like Wang, DEC, Compaq, and so many others.

I was in Charlotte NC yesterday visiting with a fast growing racing technology company that we provide service and support to for their NetApp filers. Even this fast growing, fast paced business needs help in controlling their NetApp costs for service support and expansion.

They love their NetApp equipment, but they get heartburn when they look at the prices NetApp wants to charge. So they called us, and they are very happy with our affordable alternative to NetApp.