NetApp - The Missing Shade of Blue

NetApp's $1Million Essay

I don't get paid by the word for what I write, and you probably don't either. Here's a rewarding way to change all that; pick up your pen and in 500 words or so you could be getting the equivalent of $2000 a word.

Yes, it's back; NetApp's $1 million dollar virtualization challenge. (You might remember the original challenge, and what we saw as a need to call out the exceptional advantages of NetApp's guarantee.)

And we're not looking for a Shakespeare.

In no fewer than 500 words, describe your current storage infrastructure and how it supports your VMware environment. Please include as many details as possible to differentiate your submission, such as: number of servers and VMs deployed, type of current storage connected to your virtual servers, your desired business metrics, ROI and timeframe for a new storage deployment in your virtualized environment.

My colleague Vaughn Stewart has made a couple of observations on his latest blog about the program, which is based on our virtualization guarantee. Vaughn makes a point about RAID-10 vs RAID-DP in the guarantee;

As many of you may know, this program has never been without its share of criticism by other storage vendors. In my opinion, I believe this criticism has been fair as a component of the guarantee program compares RAID-10 and RAID-DP on the merits of storage efficiency. While both technologies are near equals in terms of performance and data protection it is unfair to compare the 50% utilization of RAID-10 to the 87.5% provided by RAID-DP

I beg to differ (and did so in my original blog). I believe it's valid to compare the two, because the alternative -- RAID-5 -- just isn't up to the job of protecting your data.

I said back then; so why not compare to RAID-6?  The truth is, we could and would, but we still don’t find our competition selling it. It’s there on the spec sheet as a solution from some vendors – not all have the capability -- but it’s rarely put forward as a viable solution.

And that's the extent of our disagreement. The rest of his blog is spot on; and with public disclosure of the results, I too think we'll see high rates of savings without including RAID-DP.

Get scribbling!

.

Add Lightness

My regular readers (5 or so) will now know that I'm the proud possessor of a Caterham 7, a direct descendant of Colin Chapman's iconic flying machine, the 50 year old Lotus 7.

Constructed from a tubular space frame chassis, clad in as little bodywork as is possible, and with small and light engines, the fundamentals of a car that was said to be “too fast to race” hasn’t changed in half a century. Chapman held that:

"Adding power makes you faster on the straights. Subtracting weight makes you faster everywhere."

He was one of the great engineering geniuses of the race and motor industry.

While doing a bit of research on cloud architectures, I stumbled across an email archived on an internal engineering wiki, sent by Dave Hitz way back in 1997. The discussion was about adding a specific feature to a NetApp product, and the upsides and the downsides of doing so. (I've edited Dave's missive for brevity.)

Hey -- I love philosophical controversies, so I'm going to dive right in!

... One of the most important questions you can ask about an appliance is: Should we add this feature?

... The potential drawbacks include:

• The feature decreases one or more of "fast, simple, reliable."
• Testing the feature (in every new release FOREVER) slows down our release cycle and hurts time-to-market.
• Documenting the new feature makes the manual thicker and harder to use.
• The feature has disadvantages that are hard to explain, and customers who try to use the feature may get burned.
• Very few customers actually need the feature, reducing benefit.
• The feature is expensive to implement...

So how are my little car's design and Hitz's email connected?

I'm not a engineer of either motor or software variety, but I see the similarities. What Hitz is identifying are the same engineering principles as espoused by Chapman, applied here to NetApp's software. As Einstein said (yes, it's rent-a-quote week here at ShadeOfBlue Towers);

"Any intelligent fool can make things bigger, more complex, and more violent. It takes a touch of genius -- and a lot of courage -- to move in the opposite direction."

When I look around at the creeping featuritis that rival storage systems employ, the scale and size of the moving parts, and the horrendous complexity of the solutions they sell, I'm sure that the competition are employing intelligent fools. They certainly show little appetite to move in the opposite direction. I'll not bore you with examples, as I'm sure you know who they are and what systems and solutions they sell.

That's what struck me about NetApp when I joined NetApp some 4 or so years ago, and strikes me now. Our solutions are no more complex than they need to be; fast, simple, reliable are still the hallmarks.

But there is one item I'd like to add to Dave's email; I believe there's an extra quality that shines thorough in all our solutions and proposals. Colin Chapman, in possibly his most famous quote, said; 

"Simplify. Then add lightness."

That's the feature I think we have in abundance; added lightness. So when you're buying storage solutions -- heck, if you're buying anything -- keep Dave's list in mind for your next RFP/RFQ/RFI, And then add in this; 

In the proposed solution, please identify where you added lightness.

---

Update 8May: Steve Klinkner has an excellent series of posts on adding lightness. I've just read them (sorry Steve, I sometimes don't have enough time to keep up!), and this is one of the proofs that we take this "fast, simple, reliable" goal very seriously indeed. Well worth a read to see the kind of thinking that goes into the development of our software.

.

Who Wants To Be A Millionaire?

When NetApp came out with Virtualization Guarantee* Program (the 50% version) I was unsurprised at the ferocious chorus of disapproval from EMC, particularly from their bloggers and various EMC pundits on IT news sites. Given that the program is optional -- no one is forcing customers to take it up -- and that Gartner, amongst others, wrote approvingly of the idea, it really didn't concern me that much, and it was to be expected as EMC didn't have one (or even want to offer one). But what has got me and others really worked up was the mischaracterisation of the offer, with some pretty whacky observations about the details.

Next, we announced an updated version of the program based on 35% savings, using other people's storage (including EMC’s) fronted by a NetApp V Series. With this offer, you'd have thought we were dumping toxic waste on your doorstep if your only source of information was an EMC blog.

But now VMware has just announced their 50% savings program. At this point, I'd thought we'd get radio silence from EMC on the subject of our offer. After all, when EMC’s subsidiary (they own 90% of VMware) announces a program that's remarkably similar in nature to a competitor's program, and EMC has just spent the last 3 months dinging a program very like it, you'd probably expect at least a red face or two.

Nope. There's no shame in Hopkinton.  

Peter O'Connor (he's a Big Cheese Down Under with NetApp) is as sick and tired of hearing EMC’s usual line on this, as is everyone else. He’s decided it's time for EMC to put up or shut up.  

And as VP of Sales in ANZ, Peter's got the moolah (a cool $1million of it) to make it happen.  

The offer is to an existing EMC customer in Australia. For up to 50TB of raw storage, including controllers, software, and services for free if they’re willing to migrate their existing VMware environment to NetApp and go public with the results. There are no strings attached to this NetApp offer. Regardless of what the result is, they get to keep the kit.

It’s our response to the EMC Guarantee, listed below.

• It's not fair to compare RAID-DP to RAID-10.
• Nobody uses RAID-10 for VMware. 
• The performance of deduped volumes is bad, bad, bad!
• It's all too hard to get working.
• EMC can do dedup for VMware too! Honest.
• It's only a marketing stunt!
• The VMware guarantee is nothing to do with us.
• blah drone blah-blah drone

We’ve challenged EMC to make the same offer, and go head to head on demonstrating their value in public in a real customer environment and compare their results with ours. Let's see what happens.

.

Cisco's UCS; NetApp's USS

Over at StorageRap, Marc Farley can't quite work out the boxes on this slide from Cisco's recent Unified Computing System (UCS) presentation;

"I couldn't figure out what they were trying to say here.  Anyone know why over provisioning would be to the right on a business agility axis?  It might make sense if it meant to imply that thin provisioning was an element of Unified Computing in the upper right sector, but I haven't seen anything about Thin Provisioning's role."

Marketing; it can be confusing stuff, especially as 3Par's only notable feature in this area is thin provisioning. Here's my take on the Cisco slide.

Unified Storage is more than just thin provisioning.

• Does one storage architecture address all your storage needs? Does all the storage run under a single operating system?
• Are all protocols supported natively in the same box?
• Are all your business needs managed by single storage architecture?
• Can you easily consolidate storage for space, power, and cooling efficiency?
• Are you forced to move to a different platform to meet objectives for performance or availability or scalability?

Where, I wonder, does that put everyone else's storage solution?

Get Thin, Stay Thin, Run Fast

I was part of a webex yesterday to a number of NetApp partners about VMware, and how we add value to the business of virtualization. Presenting along with me was Mike Shea, who made me aware of something he calls the Storage Tax. I hadn't thought of it this way before.

Storage Tax

Mike's take on this is pretty straightforward. Virtualization of servers and applications is meant to bring savings; significant ones. Is it unreasonable to expect savings of 15 to 1 or even higher? Perhaps, perhaps not. Let's agree on a lower figure; 10 to 1. In fact, the figure doesn't really matter, because this is all about fractions, so make your target what you wish.

Here's the list of taxes you pay when you start down this path.

First, the shared storage tax. You need a shared storage system, SAN or NAS. They are (unfortunately!) not free. You loose part of your savings.

Second, the backup and recovery tax. Actually, more the recovery tax; backups are precious, restores are priceless in a VM environment. There's a new cost you didn't have before.

Third, more staff. You might have made your infastructure virtual, and shrunk the physical. There extra virtual objects need managing; VM savvy staff that manage more LUNs, more VMDKs, more objects you didn't have before. You can't make your staff virtual (yet!).

Fourth, tiered licensing all round for the software, from the VMs to the storage.

You can probably think of more storage taxes, and by the time you've finished adding them all up, a 2 to 1 saving might seem reasonable. That is, as long as you haven't told your finance people that you'll save them wodges of wonga, and all you come back with is a few small handfuls of lose change. It's a long way from the 10 to 1 you promised them.

That's where NetApp comes in. We can get back some of those taxes, and here I'll discuss just a few of the technologies that can help push the ratios back up towards your original goal, and point at people who really know this stuff back to front.

Get Thin

Going the VM route is about thinning out your servers, and getting better utilization of what's left. Getting thin with your storage is equally important. NetApp makes this easy; because the storage pool is virtualized in a NetApp SAN or NAS system, we can clone very thin copies of object like VMware images. We have a blog that focuses on virtualization (bookmark it now!) and a recent post had this;

"An excellent example of low cost operation is captured in this video where we host 5,440 VDI clients in 10 GB of storage and just 32 Volumes."

Stay Thin

OK, so we've driven down the storage tax of commissioning of new VMs. To stay thin is another trick we can employ, and for that, NetApp offers deduplication. Simply put, having shed those ugly pounds of storage fat in the first place, keeping those pounds off becomes essential, otherwise the storage tax just balloons out of control. 

How thin can we keep it? YMMV (Your Mileage May Vary), but here's another virtualisation blog (bookmark it now!) where customers regularly get 50 to 80% savings through deduplication.

“NetApp has saved us a lot on storage for VMware. When we originally set up VMware [at Duke Institute for Genome Sciences] , I allocated about 2.4TB for it. With NetApp deduplication, I’ve been able to shorten that down to less than 700GB. We now see an average of 83% reductions in redundant data on our VMware system.”

Run Fast

One other advantage to a NetApp system is being able to perform well, even with 1000s of VDI images. Deduplication not only reduces the space footprint, but allows performance improvements by keeping deduplicated data in a lookaside cache. Another of NetApp's bloggers (bookmark this one too!) discusses NetApp's PAM technology and how it relates to VMware;

Today, we've released yet another VMware related Technical Report (TR-3705) titled "NetApp and VMware VDI Best Practices" that addresses capacity and performance requirements, sizing techniques, system configuration validations, VDI boot storms, Login storms and the potential effect on storage as well as the role of the Performance Acceleration Module (PAM) as a secondary cache in VDI environments.

Storage Tax Reduction

NetApp provides more -- a lot more -- to reduce the storage tax and get back towards those 10 to 1 savings. And I'm grateful to Mike Shea for making it so simple to understand!

Some of those links again;

• The Virtualization Effect: NetApp's Virtual Team blog
• Nick Triantos; Storage Nuts & Bolts
• Vaughn Stewart: The Virtual Storage Guy

All worth a read.

---

[For my 7 regular readers, apologies for the long gap between this post and the last. It's been hectic here at ShadeOfBlue Towers.]

.

Squeeze Until the Pips Squeak

I’ve been blogging on storage efficiency for a while now, and it’s pretty apparent that NetApp have a great range of technologies for getting more out of a storage purchase than just about anyone else.

But what about existing customers? How do they get to make the pips squeak?

Chuck Hollis makes the good point that you really can’t just depend on providing better stuff for only new customers in the current economic climate. There’s a need to reach out to your existing customers and make sure they’re getting the best value from their storage, regardless of when they bought it.

NetApp has been doing this for a long time with our Storage Availability Audit (SAA). It’s built on the AutoSupport data that we collect from every system on support.

Performed quarterly (premium support) and semi-annually (standard support), this regular health audit offers a review of installed NetApp equipment and is followed by delivering the report to the customer.

It’s a pretty holistic process (no pun intended Chuck), covering a wide range of storage management practices. A sample of a small part of this in-depth report is shown below.

• An executive summary with recommendations for action based on NetApp best practice information
• System summary and technical summary
• Trend analysis that visually represents the peaks and lows of availability, capacity, and utilization
• Capacity summary information
• Expanded performance summary information
• System availability information
• At-a-glance section with system, component, software, RMA, alerts, and availability information
• OS/firmware summary and details

If you’re an existing customer, you can get the best out of this and other AutoSupport services through the NOW website. Start squeezing.

.

Keynote at Insight

It’s been a long week here in LA at our SE Insight conference, but a good one. Old friends, new contacts. As was the case last year, we had a large contingent of partners, and the second conference in Prague will have even more in attendance than here if I’ve read the numbers correctly.

Plus some really great sessions with plenty of information. Of note was this year’s keynote, which featured the two Toms. Tom Georgens, our President and COO, and Vice Chairman Tom Mendoza (who was, as usual, outrageously funny while making a serious point).

What was most interesting to me though was Tom Georgen’s talk, and its focus.

Tom mentioned the arguments that had erupted around the issue of various storage system capacity “overheads”, imagined or otherwise, in the blogosphere. Chuck Hollis of EMC fired the first salvo with some distinctly dodgy research, lots of people including me got stuck right in and pointed out where EMC’s calculations were flawed and based on outdated information, heat and light were generated in varying proportions and eventually a truce was declared. Of sorts.

But why would the COO of NetApp even see this blog spat as important, far less mention it?

Tom was making the point that in an economic downturn, everyone -- but everyone -- wants to minimize spend and maximize efficiency. It’s all about storage efficiency. Unless you’ve been living in a cave for the last six months, the reason it’s the right message at the right time should be self evident.

It’s not the first time he’s spoken about this. At this year’s Oracle World keynote in September his speech was Do More, Store Less: Efficiency Gains and Breakthroughs for Your Business, and on our latest earnings call it got several mentions.

Deduplication for primary storage (and VTL), thin provisioning with FlexClones, RAID-DP and more that allow us to make a 50% space saving guarantee with VMware. We can do that because cool NetApp technologies address the hot topics that customers care about. And this one, as Tom pointed out, is red hot.

.

Hairy Canary over SSD with 64 Bits of NFS

Chuck’s latest hairy canary against NetApp deserves correction. I thought about commenting on his blog, but there are so many points to make, I decided to address his “complaint” in a little more depth over here.

64 Bit – Over A Decade Ago

What’s the deal with x64 multicore technology? Just what might qualify NetApp to "lead the 64 bit discussion"? Here’s what.

Data ONTAP is 64 bit and endian clean. In fact, we were running on 64 bit systems with the NetApp F540 in 1996.

With the CX4 just recently getting a 64 bit OS – October 2008, just this year – EMC is playing catch up; they’re over a decade behind.

That’s why NetApp, and not EMC, is leading.

Solid State Storage

NetApp are getting criticized by Chuck (again!), this time for making public our SSD roadmap. A roadmap that’s freely available to every customer of NetApp could be considered more of a public commitment than those given under NDA (Non-Disclosure Agreement). Where’s the downside? Do customers want to know or not? I think they do.

That leads on to another Chuck complaint. Just what might qualify NetApp to "lead the SSD discussion"? Here’s what.

Dr. Jon Elerath, reliability engineer at NetApp, will represent the International Disk Drive Equipment and Materials Association (IDEMA) and present at the JEDEC Board of Directors Meeting on November 3, 2008. His presentation is titled “How IDEMA Can Help the Development of Reliability Standards for Solid-State Storage.”

Elerath is the chairman of the IDEMA Reliability Standards Committee, which is comprised of members from all the top storage producers (including Seagate, Hitachi GST, Western Digital, SanDisk, and STEC) and users (including NetApp, HP, Sun, Intel, SanDisk, IBM, and SNIA).

We did much the same in May 2008 too.

And NetApp is a Founding Member of the SNIA Solid State Storage Initiative. Val Bercovici, Office of the CTO, NetApp, was also elected as the inaugural vice-chair of the SSSI‘s Governing Board.

That’s why NetApp, and not EMC, is leading.

NFS Industry Leadership

Capping it off, Chuck complains about our leadership of pNFS. Just what might qualify NetApp to "lead the NFS discussion"? Here’s what.

NetApp has extensive participation in the IETF, particularly with regard to NFSv4. More specifically NetApp representatives participate as:

• Co-chair the NFSv4 working group
• Co-author of the NFSv4 specification
• Document editors for the NFSv4.1 specification
• Spec writers for NFSv4.1 (including pNFS)

Brian Pawlowski, NetApp’s CTO, co-authored RFC1813 (the NFSv3 specification) in 1994 and is co-chair of the NFSv4 working group.

Michael Eisler. Co-authored RFC3530 (the NFSv4 specification) along with Dave Noveck, another NetApp employee, and Michael is currently an editor for the NFSv4.1 specification. Take a look at his blog and his biography; it’s wall to wall pNFS.

Trond Myklebust of NetApp is the Linux NFS maintainer (including pNFS), with a hugely significant contribution over the years,

I’ve probably missed out several other major NetApp contributions and contributors to NFS. My apologies to everyone; there’s so much that we do.

NFS has been a pretty pivotal standard for the industry for decades. NetApp is deeply committed to it, and has been for over 14 years.

That’s why NetApp, and not EMC, is leading.

.