NetApp hits back at Wikibon in cluster fluster bunfight The Wikibon consultancy has published an analysis saying that the move from NetApp's ONTAP 7-mode to clustered mode (cDOT) wasn't worth it, and suggesting ONTAP was not a great choice in several application areas. NetApp's Lee Caswell, VP product, solutions and services marketing, sent us the following riposte. The Wikibon …
Hi All, Dimitris from NetApp here.
It is important to note that Mr. Floyer’s entire analysis is based on certain very flawed assumptions. Here is some more detail on just a couple of the assumptions:
1. Time/cost of migrations:
a. The migration effort is far smaller than is stated in the article. NetApp has a tool (7MTT) that dramatically helps with automation, migration speed and complexity.
b. It is important to note that moving from 7-mode to a competitor would not have the luxury of using the 7MTT tool and would, indeed, result in an expensive, laborious move (to a less functional and/or stable product).
c. With ONTAP 8.3.2, we are bringing to the market Copy Free Transition (CFT). Which does what the name suggests: It converts disk pools from 7-mode to cDOT without any data movement. This dramatically cuts the cost and time of conversions even more (we are talking about only a few hours to convert a massively large system).
d. NetApp competitors typically expect a complete forklift migration every 3-4 years, which would increase the TCO! Mr. Floyer should factor an extra refresh cycle in his calculations…
2. Low Latency Performance:
a. AFF (All-Flash FAS) with ONTAP 8.3.0 and up is massively faster than 7-mode or even cDOT with older ONTAP releases. To the order of up to 3-4x lower latency. cDOT running 8.3.0+ has been extensively optimized for flash.
b. As a result, sub-ms response times can be achieved with AFF. Yet Mr. Floyer’s article states ONTAP is not a proper vehicle for low latency applications and instead recommends competing platforms that in real life don’t perform consistently at sub-ms response times (in fact we beat those competitors in bakeoffs regularly).
c. AFF has an audited, published SPC-1 result using 8.3.0 code, showing extremely impressive, consistent, low latency performance for a tough workload that’s over 60% writes! See here for a comparative analysis: http://bit.ly/1EhAivY (and with 8.3.2, performance is significantly better than 8.3.0).
So what happens to Mr. Floyer's analysis once the cost and performance arguments are defeated?
Thx
D
Hi D, a few questions, as your responses don't jive with administrators I've talked to.
1 a) Everyone I've talked to has said CDOT Migrations are a miserable bitch and usually edge towards "unmitigated disaster". Can you explain why this is? What are Netapp admins in the feild doing wrong? Why are their reports different from your own?
1 b) GTFOing NetApp is a one-time pain, followed by not having to deal with that shit again.
1 c) What are the risks with your CFT? DO you offer guarantees in the form of gigantic piles of cash if the CFT migration blows up and craters a company for hours/days/forever? How much cash, exactly?
1 d) I present to you: Pure Storage, amongst many, many others. Forklift upgrades are retro.
2 a) Congratulations: you made an array go faster by putting flash in it. Do you want a lollipop? How does it compare to competitor hybrid or AFA arrays on a $/GB/IOPS or $/GB/Latency basis? In case you missed it, that's the bit that matters. Your ability to compete with yourself is not relevant.
2 b) This is an interesting claim, as what I hear is that it is NetApp who can't achieve consistency of storage response for low-latency (or high throughput!) operations. I - and everyone else - would love to see these bake-off numbers, including details of the workload and the competitors against which it was measured, whether or not this was in a mixed workload environment, and more. He says/She says. Either could be right here.
2 c) This is a start, but it is also just one benchmark. See above.
Now, D, I am emphatically NOT saying that Netapp is wrong and Floyer is right. I am entirely willing to entertain that either side of this is wholly or partially right. What I am saying is that Floyer's take is backed by a large amount of anecdotal evidence that seems to reflect popular opinion on the subject.
I do not thing "blogs at dawn" is the answer. It's time for some serious engagement with independent testing groups. There are many. Pick a few and let's put this to bed so we can all focus on why NetApp doesn't have a proper hyperconverged solution. :) (No, EVO:Anything doesn't count. Look at your sales figures, you know it to be true.)
1a - Anecdotal evidence is interesting, especially when colored with confirmation bias. I assume you've spent your days talking to every single storage admin that's ever migrated to cDOT?
1b - Where is your empirical evidence and precise examples of where dealing with cDOT/ONTAP is so much worse than any other vendor, so much so that it's worth the *extra* downtime to do so? Because all I see here is unfounded opinion.
1c - I'm fairly certain that there are legal avenues that can be taken if "CFT craters" on you. Luckily, CFT has rollback methodology built in, backed by decades of proven SnapShot technology.
1d - I agree - Forklift upgrades are retro, which is precisely why customers are moving to cDOT.
2a - Wait. So first it was "NetApp was slow to the flash market" and now it's "congrats you have an all-flash solution. Big deal." Which is it? And I'm sure you realize that AFF is *not* classic FAS, right?
2b - The data is out there and more is coming.
2c - You complained about no proof in 2b, then shrug off the proof in 2c. This makes little sense.
I will say that your entire post would have been perfect if summed up in your last two paragraphs.
There *does* need to be industry-wide standards and testing done by independent groups. It needs to be impartial. The FUD slinging has to stop, and honestly, it starts with the people with the largest mouthpieces, like The Register.
Hi, AC. Let me respond to your epistles point by each.
1 a) "Anecdotal evidence is interesting, especially when colored with confirmation bias. I assume you've spent your days talking to every single storage admin that's ever migrated to cDOT?"
No, I haven't talked to every single storage admin that's ever migrated to cDOT. I have talked to all those I could find and who were willing to talk to me. Just like I talk to sysadmins about their experiences with any and every bit of infrastructure (especially storage) or software they are willing to share experiences on. The more I learn about pain points the more informed my questions can be when I get the chance to talk to vendors. That's, you know, my job.
Regarding confirmation bias: I don't have any on this topic. Yes, I think NetApp is probably doomed, but the reasons for that are more related to issues around the support experience, partner/channel difficulties, sales targets, marketing and the utter incomprehension on NetApp's behalf of community management.
The cDOT thing is actually fairly incidental in my calculations. Despite that, the issue has been raised so many times, by so many sysadmins that when I saw what appeared to be a NetApp employee adding info to an article I took the opportunity to ask questions in the comments forum instead of using my full blown article pulpit.
The comments section receives less than 1% of the readers of a main article. NetApp has declined to assign me a PR flak to whom I can ask questions when I have them. This seemed a perfectly valid way to ask questions in a manner that would be relatively low impact to NetApp itself.
1b) "Where is your empirical evidence and precise examples of where dealing with cDOT/ONTAP is so much worse than any other vendor, so much so that it's worth the *extra* downtime to do so? Because all I see here is unfounded opinion."
Well, were I going to take the time to ask my sources to get official permission to discuss this, be named, and document their issues in full I would damned well expect to get paid for the effort. This means writing up a full article. Given the tales told around the boozer about this that article would be damning. Bordering on assassination. I have no interest in assassinating NetApp. At least not over the cDOT thing.
As to whether or not it is "worth the [extra] downtime" to migrate away from NetApp to another solution, different people have expressed different opinions on the topic. For some, they have heard of so many things going wrong during cDOT migrations that they flat out do not trust NetApp's claims of seamless migration and wouldn't try it, no matter what the spokesdroids say. For those customers the question then does NOT revolve around downtime (or lack thereof) but instead around the ROI and TCO of the different offerings.
Others are using solutions such as Datacore or Falconstore to migrate workloads from NetApp to other solutions live and without interruption. Some competing storage vendors have other tools which make migrations easier. Migrating workloads off of one storage vendor's tin and over to another its own industry.
Others might be willing to trust in NetApp's cDOT migration capabilities but view the cost of staying with NetApp as being higher than a competitor + the outage/effort required to migrate. And yes, there are those who have - successfully and unsuccessfully - felt that sticking with NetApp and doing the migration is the best path.
I am not claiming - nor have I claimed - that any of these paths is "correct". I asked questions and I was hoping for a reasoned response that would allow me to gather more data on the topic.
1c) "I'm fairly certain that there are legal avenues that can be taken if "CFT craters" on you. Luckily, CFT has rollback methodology built in, backed by decades of proven SnapShot technology."
What are those avenues? What is covered? Under what circumstances? Vague promises are irrelevant here. As for "has rollback technology"...great? I mean, that's some comfort, but the risks involved are so high that the existence of one possible technological remedy is just not remotely relevant compared to the importance of the financial and legal remedies that are available. As you seem to be involved in teh technology industry, I am hoping you are more than passingly familiar with risk management.
1 d) "I agree - Forklift upgrades are retro, which is precisely why customers are moving to cDOT."
I can see this as an easy enough marketing message. I am trying to match the corporate bravado with actual reports from systems administrators. To date, it would seem that those who trust in this particular bit of "messaging" are vastly outnumbered by those who do not.
Seem is the operative word here. I have stated what I see on the ground. I welcome rebuttal with facts and numbers that can be verified.
2a) "Wait. So first it was "NetApp was slow to the flash market" and now it's "congrats you have an all-flash solution. Big deal." Which is it? And I'm sure you realize that AFF is *not* classic FAS, right?"
First off, I don't recall saying NetApp was late to the flash party. It was, but that isn't something I consider relevant to today's systems administrators. Today, hybrid flash and all flash are common. No big deal. The question isn't "can you do it" - Synology can do it! - but "does your solution suck less than others and/or cost less than others?".
AFF may not be classic FAS, but they are really not so very far apart. While I have not personally had the opportunity to run AFF through the wringer, others whom I trust to be very good at testing these things found it to be decent on performance but less than middling on value. As always, more data to better refine analysis is better.
2b) "The data is out there and more is coming."
Hyperlinks would be appreciated. What I've seen so far does not convince me that NetApp is the be-all and end-all of storage by a long shot.
2c - "You complained about no proof in 2b, then shrug off the proof in 2c. This makes little sense."
Um, no. I acknowledged that the data provided was an important data point, but do not find that it is remotely adequate enough to call it for NetApp. Performance, looks great for one specific benchmark. NetApp should be proud, but that is really only one data point amongst the many needed. It also does not address TCO or ROI, nor the intersection of those two with performance.
Proper analysis considered multiple use cases, multiple (and mixed) workloads, ROI, TCO, performance, support, migration, reliability, insurance coverage, ecosystem, future proofing and migration friction/lock-in.
"I will say that your entire post would have been perfect if summed up in your last two paragraphs."
I am sure you would, as your pro-NetApp bias is pretty blatant. Which, to my mind, is aught but a stronger incentive to ask more - and more probing - questions.
"There *does* need to be industry-wide standards and testing done by independent groups. It needs to be impartial. The FUD slinging has to stop, and honestly, it starts with the people with the largest mouthpieces, like The Register."
See 1 a). I'm doing my job. What's yours, exactly?
And while you're at it, please have the bravery to use your real name. If you want to go after me, The Register and/or anyone else, don't hide behind the coward's veil.
Ding Ding Ding we have a winner. Mr Pott.
Reality check beyond marketing bushwackery
1) The majority of netapp customers do not get to perform their own data migrations because it is indeed a painful process
2) If it were that easy netapp wouldn't have had dedicated transitions teams. Transition is code word for "migration"
3) Copy Free Transition is painful too.
a) It requires building a new cluster
b) has specific 7mode ONTAP requirements which address a subset of netapp's install base
c) Apparently it's not guaranteed because there's a rollback process. If it fails then what?
4) Will keep out of commission for 4-8 hours. That's when you're are DOWN.
You can read all about it here before it 404s...
While you read this, think if a customer can actually do that on their own.
https://library.netapp.com/ecm/ecm_download_file/ECMLP2346396
I agree, Wikibon are a bunch of paid for Tier 3 self-serving individuals on EMC's and Oracle's pocket but you shouldn't complaining because this is how the game works with the so called "independent" analysts. It appears you too have had some success with them recently.
My advice is:
1) Stop the stock buybacks
2) Invest in the business
3) Acquire worthy products
4) Stop the bleeding and the perennial layoffs
5) Acquire competent upper management
6) Stop celebrating for no reason
Become competitive again because now the only people that pay attention to you are those who have invested their entire careers in ONTAP and are set in their own ways. The rest have moved on. It's in your numbers.
Goodbye and Good luck
The No Spin Zone
Couldn't agree more. A move to cDOT is complex and hard. If you get stuck halfway through the "migration" you are really screwed.
Freedom of choice in the Cloud - comes at the cost of being locked in to cDOT. So not really freedom.
The decision to shove cDOT down the (existing) customers' throats was a huge gamble that didn't pay out.
Trying to make customer's pay for "cDOT migration services" wasn't a smart idea. Who wants to pay to be the guinea pig ?
NetApp Support wasn't ready when cDOT came out burning early adopters.
Changing product names several times didn't help either.
The most Successful Netapp Reseller Partners are the ones that are heavily invested in Technical Skills. Enabling new Reseller Partners is extremely hard because they struggle with the technical learning curve and the established Partners just take their business.
Resellers prefer a (reasonably) low complexity product and to be profitable (i.e. not burning PS hours to tweak ONTAP).
"Competent Upper Management" - In the country where I worked at NTAP Managers just hired their drinking buddies; they've actually forgotten how to manage people. They did however become experts at "managing people out of the door" in the most creative ways. HR wouldn't dare to say anything.
I think NetApp is in the process of becoming slim, cheap and attractive as potential acquisition target.
It's really sad that the culture that made NetApp such a great place - will also be its downfall.
I'm going to assume that the "No Spin Zone" you referenced started AFTER you typed it. Because everything prior was basically a 12-6 curveball.
I'll also assume you're just innocently misinformed rather than intentionally misleading.
Allow me to address this point by point.
1) Migrations are not a matter of "customers not being able to choose to migrate." NetApp wants PS involved because migrations aren't just about moving data - they're about planning and fixing infrastructure issues that have nothing to do with the storage itself.
2) Dedicated transition teams were necessary prior to recent history. Now, transitioning to cDOT is getting easier and transition teams won't be necessary. And you're splitting hairs by spinning transition into "migration." Call it whatever you want.
3) CFT is painful?
a) A new set of heads is needed regardless of if it's 7-mode or a competitor system. Why is having a new cluster any more "painful" than other migration efforts?
b) There are specific 7-Mode requirements. Why is that a problem? No one is hiding that fact. And many of the requirements are a result of either the improved cDOT architecture or the fact that this is v1 of CFT. Future versions of CFT will have more flexible 7-Mode requirements.
Did the first iPhone have Siri or even have the ability to surf the web while you are on a call?
c) Is any software upgrade guaranteed? Doesn't every upgrade have a rollback/downgrade process? How does the fact that CFT have a rollback make it "painful"? It's there if you need it - it would be irresponsible to provide a migration tool that didn't have a back out method.
4) 4-8 hours of downtime is the guideline. But time that includes cabling your shelves (which could take anywhere from minutes to hours, depending on your datacenter). The actual conversion of the data takes very little time. We had an ACTUAL customer migrate 100TB of data in less than 2 hours of downtime.
How long do you think copying 100TB from a 7-Mode system to a competitor would take?
You linked to an official NetApp doc. They're not hiding anything from you. It's all out there, in multiple locations. And no, it won't "404" on you since it's THE OFFICIAL DOCUMENTATION.
As for your opinions on how NetApp should transform it's business, I can't argue with those; they're your opinions and are the only things in your post that didn't need addressing.
Hi Trevor, my response was too long to just paste, it merited a whole new blog entry:
http://recoverymonkey.org/2016/02/05/7-mode-to-clustered-ontap-transition/
Regarding your performance questions: Not sure if you are aware but posting bakeoff numbers between competitors is actually illegal and in violation of EULAs.
We have to rely on audited benchmarks like SPC-1, hopefully more of the startups will participate in the future.
I suggest you read up on that benchmark, it's extremely intensive, and we show really good latency stability.
Though gaming SPC-1 is harder than with other benchmarks, it too can be gamed. In my blog I explain how to interpret the results. Typically, if the used data:RAM ratio is too low, something is up.
Which explains a certain insane number from a vendor running the benchmark on a single Windows server... :)
Take care folks
D
Dimitris,
It's great to see someone from a vendor here, thanks for being part of the discussion... it makes this a little more human for many of us.
I have recently moved some customers away from NetApp. I more than likely won't be using NetApp anymore in the future either. This isn't because you have a bad product. It's an excellent SAN product, the problem is that it is in fact SAN. As a NAS, I simply don't see any logical reason to depend on SAN products when NAS has always been handled better by servers than storage devices. My customers range from between single shelf unit sized through US government three letter organizations with purchasing budgets of $500 million per project. I just re-architected one of those deployments to use SAN for boot drives only and that was NetApp, but we moved from what was likely to be a 4 petabyte SAN to a 16 terabyte SAN and 4 petabyte server cluster instead.
Here's the deal. 7-mode was amazing. C-mode, not so much. NetApp's documentation is far too messy and the operations of C-mode is far too complex for what the customer's need. This is 2016 and the idea of having a team of full-time storage admins seems utterly ridiculous. In my experience, NetApp C-mode is utterly unusable without advanced training which is very expensive and generally not particularly good.
C-mode documentation is pretty awful. The web gui is unusuable in most cases as NetApp support keeps saying "Oh... I only use command line...". The PowerShell API for C-mode works... barely... which means it's not manageable through either Azure Pack or System Center Orchestrator. The network management is extremely monolithic and although there's support for things like virtual interfaces, they don't scale very well past a certain point.
Redundancy is a problem too. NetApp's drive prices are so outrageously high, I often wonder why you don't just give away that racks and controllers and charge a fortune for disks. When installing a drive requires installing 4 since you need redundancy within a single array as well as redundancy in the other data center, using disks that are generally twice the average industry cost isn't going to work.
Due to licensing issues with OnTap, the resale value of NetApp devices is virtually zero. You've locked the users so tightly into your cloud management system that as a company is expanding, they know they'll have to simply eat the loss on their SAN from you because they can't effectively expect to be able to resell it. Therefore, once you buy a NetApp, you scale it as best you can until you throw it away and buy a bigger one with almost complete loss of initial investment.
Performance is a real problem as well. C-Mode has a hard limitation of 8 controllers which isn't too bad, but the expandability of each controller is really limited. There isn't much room for upgrading RAM. The big solution I just specced out contained 8 servers each per data center (across 5 data centers). Each server internally contains 52 8TB drives internally. In addition, they contain 6TB of PCIe SSD and four, two-port 40gig ethernet adapters for a theoretical 240Gb/sec bandwidth per server scaled out. We have a great deal of room to scale up as well. Each server contains 6TB of RAM too. The entire solution runs on Windows Storage Server and by the time we deploy will have a full documented RESTful API as well as extremely extensive management tools. Using scale out in the lab, we're keeping 240Gb/sec per server pretty solidly saturated at all times.
On the small end, using systems like Dell's cloud servers or Cisco's M-series servers, we can easily have insanely high performance storage for a minor fraction of a similar solution from NetApp.
So, if we're trying to move from 7-mode to C-mode, it's not really worth the effort. We can't simply switch because we can't risk our data. So converting isn't an option, it's simply a matter of buying a second netapp solution to move to progressively.Windows Storage Spaces just performs better and also is well understood by the server guys which makes it much less expensive as it means you do't have to train and employ an entirely separate team to manage storage. It's much easier to manage due to excellent documentation and integration with PowerShell and System Center.
We don't really need iSCSI except to boot blades, and frankly, Starwind is almost ready for the big time and DataCore has that covered pretty well. They both have reputations and experience exceeding NetApp's own.
Don't feel bad, compared to EMC, you're doing great... but scale out file servers on commodity hardware is the long term solution. I'm sure you're not going anywhere though... people will keep buying your stuff for years because "That's the way we always did it." but for new data centers and new deployments, I don't see the value.
* WAFL is ideal for Flash. Solidfire is for webscale only *cough Cough*
* Migrations times arent bad. it took us 5 years to come up with a mash of tools to migrate.
* Cluster mode is sooo fast. NetApp has and will always be CPU bound. You have to be a Jedi to turn this complex animal. if all else fails blame the customer and call in NetApp PS to "tune'
* AFF is NetApp's fastest growing product ever. It is also the only product NetApp have released that wasnt acquired tech in a long time. It also should be noted that its been butchering their install base to get fix the problem with disk centric WAFL. lets add a dedicated Flash tier, add in a ton of caveats and look over there we bought solidfire
i think Lee and Matt Watts need to step outside take a deep breath and continue to watch the world go by
This post has been deleted by its author
<sigh>
Flash makes everything CPU and pipe bound. If anything, CPU speed is becoming commoditized very rapidly... :)
And yes, cDOT 8.3.0 and up IS seriously fast. We have audited benchmarks on this, not just anonymous opinion. I understand that competitors don't like/can't come to terms with this. Such is the way the cookie crumbles. Look up the term "confirmation bias".
Regarding SolidFire: It's not about speed vs cDOT. As a platform, SolidFire is very differentiated not just vs cDOT but also vs the rest of the AFA competition.
The SolidFire value prop is very different than ONTAP and performance isn't one of the differentiators.
What are some of the SolidFire differentiators:
- very nicely implemented QoS system
- very easy to scale granularly
- each node can be a different speed/size, no need to keep the system homogeneous
- ridiculously easy to use at scale
- little performance impact regardless of what fails (even an entire node with 10x SSD failing at once)
- the ability to run the SolidFire code on certified customer-provided servers
- great OpenStack integration
All in all, a very nice addition to the portfolio. For most customers it will be very clear which of the three NetApp AFA platforms they want to settle on.
Thx
D
I don't work for NetApp and don't particularly care for NetApp's ONTAP strategy. That said...
WHO CARES WHAT WIKIBON THINKS ABOUT ANYTHING?
They are a tier 3 analyst firm trying to make headlines. They are desperate for attention. They cater and stroke the vendors that pay them (like IBM, EMC, Oracle, etc). They are no different than any other tier 3 pay-to-play analyst firm.
The only thing I take away from this is that NetApp clearly didn't pay Wikibon enough money.
I care what Wikibon thinks. I can point you to thousands of sysadmins and CIOs from companies ranging from 20 man to 2000 man who care what Wikibon thinks. Most of those folks care a lot more what Wikibon thinks than Gartner, Forrester, 451 or IDC.
If you want "pay to play", let's have a talk about the "tier 1" analysts. Indeed, given that I work with so many companies and have seen this from all angles, I would love to have a very public, very open, very transparent discussion about the state of marketing and analysis in tech. As a strong believer in the importance of independence, actual testing, proper research and full disclosure, I'll gladly use every single one of my pulpits to engage with that discussion.
So please, let's talk! Especially if you can get the folks at NetApp to reveal whom they have paid, how much, and for what. From marketing to analysis, whitepapers to "pay to play" let's get it all out in the open, shall we?
Wikibon absolutely get paid to do analysis. But I am more than willing to believe they are objective and relatively independent than Gartner and their ilk. Do you have any idea how hard it is to maintain some semblance of independence when analysis is your job? Let alone rake in millions upon millions?
So, hey, if your folks at NetApp are willing to open their kimono, I am more than willing to bang the drum about how the sausage is made.
What say you...are you game to put up?
"All Flash FAS (AFF) platform, which cDOT underpins, one of the fastest growing products in NetApp history."
1. It's the Flash underpinning cDOT.
2. Fastest growing as compared to: GX, Flaxhray, OntapEdge....
Has anybody had a look at NetApp's share price ? Why are excs leaving ? Lots of them...
Doesn't matter what Wikibon or NetApp says - if the customer won't buy the CDot (Betamax) OS, then Netapp is done....
Execs are leaving across the industry, for a variety of reasons. Most of the time to try to cash in on one of the 8 billion startups that are going to IPO.
What storage share price is "up" right now?
Pure's at $13
Nimble's at $6
EMC's at $24
NetApp's at $21
If you're basing a company's future on share price and exec movement in a saturated industry, you're looking at the wrong things.
Uh...you just listed share prices without remotely looking at growth, number of shares, market capitalization or a squillion other things. Share prices mean less than comparing apples and sidewalks to the inky void of space. The prices need rather a lot of additional context to tell you anything at all about the companies.
I think you just rendered every post you've made in this thread a joke, as written by someone who doesn't understand what the nether fnord they're on about, mate.
Yes, I listed share prices in response to the poster who said "Has anybody had a look at NetApp's share price ? Why are excs leaving ? Lots of them..."
And then I said "If you're basing a company's future on share price and exec movement in a saturated industry, you're looking at the wrong things."
So, your reading comprehension suggests that every reply you've made in this thread is a joke, as written by someone who doesn't understand what the nether fnord they're on about, mate.
Its not just tech stock which are down its the whole market driven by matters much larger than tech alone - think oil/China/struggling BRIC economies etc. So whilst shares prices are a measure, they are hardly the greatest measure. As the small print always says, stocks can go down as well as up.
The thing that really matters are the underlying figures and with this fact alone, NetApp is struggling, and the sales figures speak for themselves. Ignore the haters, the analysts and the fan club and the people who work for NetApp who are paid to defend it and look at the cold hard numbers. NetApp hasnt grown for years and there are no reasons that I can see to suspect that this will change.
I think this reason is manifold.
Firstly, it took its eye off the ball by being wedded to a technology it acquired some time ago which it just couldnt get to work and whilst cDOT might now be the almost finished article, times have changed since then and the market has moved on.
Despite what people think, a seamless migration to cDOT from 7Mode isnt. This 'break' makes admins think about the market. Lets assume that a migration to cDOT has the same pain as a migration elsewhere then this presents an opportunity.
NetApp is complex, simple, compared to others. Perhaps they will say that this is because it has lot of functionality and this may be true but lot of apps can now pick up the slack and provide much of this natively. What admins want in my experience is fast, simple, robust storage. NetApp may be robust and in some configs, fast, but simple it isnt.
Solidfire is confusing. One major benefit of SF is QoS for service providers. Its touted everywhere. Apparently FAS can do QoS in cDOT and has been trying to compete with SF for some time. Now they have bought it. This isnt going to grow the business, this will just plug one small potential leak in service providers where NetApp claimed to be a leader.
The competition is smart, very smart. Nimble, Tegile, Tintri, Nutanix, Simplivity etc and others are taking NetApp apart, taking their customers and beating them to new wins. There are just too many that are taking customers and stopping growth, and growth from market share increase is what's needed. Organic growth in the double digits of the last decade are just that, a thing of the past.
CIFS is no longer required. There's a reason FAS has the filer nickname and that is because it started as a file server consolidation tool before it expanded to include block protocols. In virtual environments you dont have physical server sprawl like yesteryear. A virtual 2012 file server is just as capable as native CIFS in 99% of cases and offers DFS/quota management and dedupe. Dont forget the sales tactic used by NetApp against its own OEM, IBN N series. They would tell you that IBM were always code revisions behind NetApp and you should just use the real thing. Do NetApp write and develop CIFS/SMB? Were NetApp wrong about N-Series all that time?
NFS will no longer be required - vVols - I will say no more.
Post warranty support costs - I will say no more.
Support experience - I will say no more.
The whole storage/infrastructure market is in turmoil right now. Dell had to acquire EMC to survive, EMC had to be acquired by Dell to survive, HP had to split in two to survive, Pure had to float to survive, Voilen may well need to be bought to survive, Nimble has to grow beyond its debts to survive, IBM has sold its tin business to make the books look better, Cisco needs to acquire a storage player to bolster its position, other startups need to float or be bought and neither of those seem palatable to either the market right now.
Cloud is relevant but is only a distraction right now. AWS and Azure are the only true competitors in this market and they dont house NetApp kit for their storage as far as I'm aware. Google will continue to erode file shares with GMail and Docs as times goes by. The rest of this 'cloudy' picture is merely a marketers wet dream and is just a term for 'IT in another place'. NetApp needs this hosting and service provider business and as my point above explained why this is no gimme.
So, will NetApp, in my opinion yes but for how long and it what form. I dont think anybody really knows but one thing is certain, it will get worse before it gets better.
I find this whole thing very interesting, and quite amusing. Re; analyst firms. Someone said wikibon is "pay-for-play" - guess what, they all are pay-for-play. Look at all the top ranking analyst firms and none of them are a non-profit.
I've never or very RARELY have read a hard line piece from an analyst firm that really gives the customers something to think about when considering a vendor. So Wikibon did, and now we see all the hand-waving from NetApp, and marketing regurgitation talking points (even though D is very technical and highly respected for his skills, his response seems to be a tailored retort). In my opinion, analyst firms should inform and educate the buying community (customers), and vendors should all grow up and take it like pros. The customer is always right, right? No, not in this case, it is the customer's fault but PS will save the day.
Get over it, NetApp. Read the piece for what it is, and prepare your sales and partners with proper response. Who knows, this may have never hit the big time without your own help. You gave it wings to fly all over the interweb, and now this little piece that even I didn't hear about until I hit one of my social media sites this morning. Wikibon wrote this on the 18th, Chris picked up on it on the 27th and stirred the bee's nest, then NetApp chimed in with a VP title to debunk, followed by a series of employee and probably NetApp partner comments to follow. Don't you wish you could have really thought this through before giving this the attention it maybe didn't deserve? someone is corp comm will be updating their linkedin profile soon.
As for the cDOT seamless migration, I can tell you from experience there are seams and they burst. Honestly, you should have fixed that very early on, but since it took you nearly a decade to deliver on cDOT to begin with, I guess you need to go to market at some point (isn't the Spinnaker stuff that you bought in 2003 part of the whole cDOT thing today? maybe I'm confused and lost, I cant keep track of your acquisitions or your intended strategy - except that OnTAP rocks and anything else is just crap).
Who is more important, your customers or proving Wikibon wrong? Seems like you are spending way too much time trying to prove Wikibon wrong and not enough time with customers and creating your strategy and vision based on who wants to buy your solution. Hey if the customer says, this sucks and I hate it, I'm thinking of going with HP, EMC, or fill in the blank. Shouldn't that be a red flag?
Customers appreciate when vendors are big enough to take the punches and work through it with them side by side. Nobody likes it when the hands start to wave, you begin to wonder where the truth really lies.
Just an observation.
Now if I could upgrade my 7-mode systems to 8.3.x and then convert it to a switchless c-mode with very minimal downtime that would be great. Though the CFT feature is a good step in the right direction and depending on how long the downtime is that may be the way I upgrade going forward.
So yes I'm a Netapp customer and at work we have a couple of C-mode clusters(Each over 8PB in size) and overall I don't have many issues with it. With C-mode I can migrate volumes between older disks and newer disks non-disruptively, move data lifs to less loaded controllers if needed and grow the system out without disruption - compare that to 7-mode.
Overall it has been very solid so really the biggest complaint is migrating from 7-mode to C-mode and not C-mode itself. With that being said there are a couple of things that give me stress with C-mode. One is that it doesn't support sync snapmirror like 7-mode and for whatever reason the developers here refuse async. That's been our hold back for migrating the rest of our systems. The other issue at least for me is setting up 'Flexarray' to connect to their e-series systems. It is a time consuming task so that is costly and the overhead that Ontap takes on top of what the e-series already took is quite a bit. Lastly the Netapp controllers can only manage lun sizes up to 6TB but the E-series could export much larger luns so I'm waiting for that to be addressed in the Flexarray code. However once it's set up it works very well and we haven't had any complaints. All in all it's been very very solid and we haven't suffered an outage or data corruption and performance has been fine for what we use it for.
I'm talking about something entirely different "Flexarray". Each disk from the E-series storage is seen as a LUN from the Netapp controller side. In Flexarray terms the max lun size is currently determined by the largest supported drive size which currently sits at 6TB. Again I'm talking about 'Flexarray" its a completely different beast than traditional luns on the Netapp using DS4243 or DS2246 storage hardware.
zbmwzm3,
At least NetApp has worked to allow you to recoup some of your investment in other vendors storage and still be able to move to NetApp.....they could let you throw all your money away and start from scratch. If your requirements for LUNs exceed 6TB, then perhaps flexarray it is not the solution for you. That doesn't make it bad, it just means you have to match up the solution to your requirements.
Even though I’m fully aware that this site is nothing more than link-bait akin to BuzzFeed or Fox News, it still amazes me to see one of the “reporters” come back with “I hear cDOT migrations are a disaster”.
Trevor, if you’re hearing this, it’s because of 1) that person has an agenda to perpetuate this myth, and/or 2) there is a level of ignorance or unwillingness to put in the effort.
Any vendor I’ve ever worked with (and I’ve worked with all of the ‘established’ vendors for the past ~15 yrs) has been willing to bend over backwards to help get the job done, regardless of the task. This is especially true for NetApp.
If you’re not able to do a data migration, which is most likely part of your job and has been a standard practice for decades, then that is your problem and not NetApp’s. There are plenty of tools and resources available to get the job done properly, smoothly and even non-disruptively in some cases. You just have to put in the effort.
If you’re a “higher up” that isn’t hands-on with the gear, 1 and 2 above apply to you also.
What gets me is all of this hating on NetApp for requiring a migration in the first place. As if they’re the first company to do so. It’s one minor detail that the competitors and haters cling on to in order to sling ‘mud’. Unfortunately, it’s effective. Until recently, this has been standard practice with almost any vendor product and still is with most. Get over it! XIV to/from DS8K? Migration. VMAX to/from VNX? Migration. Hell, our upgrade to VNX2 from VNX1 required a migration. We don’t have an XtremIO, but I heard there was a disruptive software upgrade that required data migration (since we’re throwing around ‘I heard” rumors).
Every time I see the “migration disaster” nonsense, I think of the adage “If you have a bad experience, you tell 10 people. If you have a good experience, you tell 1” (or whatever numbers you want to use). If these bad experience you speak of are in fact genuine, then of course they’re going to tell as many people as they can. When you hear this, see 1 and 2 above.
We’ve already been through a hardware refresh after we moved to cDOT and it required absolutely no downtime for our applications. Coordinating downtime with the application owners used to be a huge consumer of our time. Not any more.
Dude,
That is such an biased argument that it can't be left standing unchallenged. Your basic point in comparing ds8k to xiv and vnx to vmax required migrations is silly. All of those examples are from untapped product lines in their respective company portfolio. In case of cdot netapp toured it as the continuation of the product line and its quite frankly bare minimum to have upgrades within the product line. The companies and products that misses that loose customers and marketshare and flatline before they decay. Once that happens the best they can hope for is an acquisition so now competent people can run the show.
I don't recall that vnx1 to vnx2 migration was next but the clariion to vnx was indeed a bit messy. Well revenue since flatlined and declined and the company is selling itself.
Netapp is flatlined, maybe even declining, uptake of the latest and greatest had been a disappointment (quoting the chief executive) and I'm sure they would love to "consider strategic options" as the euphemism for sell off is called.
Modern predicts don't have outages; software versions and hardware updates are online while in production. Companies that can't deliver that and more are desperately clinging on to install base and financial engineering to keep customers happy (hello emc, flatlined yet profitability is down 27% -I guess the current offerings are of less value to customers than the old offerings).
I'm a of the opinion that these companies aren't just f*d by the cloud, they are also f*d by every upstart delivering products with the experience you get from cloud; simplicity, ease of use, always on, zero outage windows. It's this double whammy that is so painful for them to experience, and its totally true that their engineers and product bosses can't comprehend what is happening in the world around them.
Netapp was a great company, and so was Sun Microsystems. Sun didn't see the transition to x86 and distributed scaled out computing and got f*d by that. Netapp is in that exact spot of denial that sun was in circa 2002.
/w
This post is old, but thought I'd add my 2 cents. NetApp like any large company has made some missteps. Development has taken longer then anyone would like for cDOT for which all that have worked with or own NetApp would agree the migration tools and support tools around cDOT are still immature which is a pain, but they are headed in the right direction, but NetApp's philosophy is still pretty sound and where they want to go is what some customers will need for time to come. I suppose the question will be can they make it there before they "lay off" or drive to quit all the folks that have made them what they have been to satisfy the Wall Street gods and foster a culture of don't rock the boat or I'll loose my job which will kill the company long before it get's sold (which may already be done). NetApp like Microsoft and Apple started with a good idea and capitalized on it by taking risks in it's early days. Buying Spinnaker was a good idea, just not capitalized on in a timely manner. cDOT was that next step and what the industry needed 10 years ago and if they could be where they are today 5 years ago, we wouldn't be having this conversation today; however we are in 2016 and the storage field has drastically changed and the options available to customers to get the job done (many cheap start ups spending a lot of venture capitalists dough to give away their wares hoping for a suitor to make them rich). NetApp got caught with their pants down milking cheap disk for all it was worth for profit sake when it should have been pushing flash, leading the industry and so upstarts have in some ways marginalized NetApp (at least for a time). NetApp really needs to execute well in the days ahead or it will find itself like EMC and since they just laid off 1200 top notch folks (not dead weight) it would seem that they have chosen to put lipstick on the pig as corporate types tend to do looking for their golden parachute while they leave everybody else high and dry with a shell of a company...so much for #1 company to work for....if you were on the support/services side of the house, you have known this for a long time. NetApp propaganda will be it's own undoing....they need to get off the cool-aid, roll up the sleeves, and get to work solving customer problems and providing solutions that meet the needs of partners/customers or standing behind them when they don't. The days of obscene profits are over in the IT industry....but there are many here that still think you can sell something for ridiculous amounts of money and just expect to walk back in the door 3 years later and do it over again with no effort....NetApp is guilty of this as is every vendor and VAR have been. Some sales people want to work 10 days a month and play golf the rest while the SEs and PSEs tear their hair out and work insane hours trying to make good on empty promises.....this needs to stop....and it will by the company going belly up. Have to give Dell credit when they went private.....everybody said they were dead, but just look who bought EMC. Something to be said for being private and making decisions that make sense for the company and not for Wall Street gods who know nothing more than a P&L statement and an Excel spreadsheet. Mr. Dell should probabally be the one running for president, but he is busy keeping his company relevant in a very complex IT landscape rather than declaring bankruptcy and calling it a day.
But let's look at the flip side of this coin too:
We have a bunch of venture capital startups slapping hardware together and loosing $2 for every $1 made....we say how can they do it so much cheaper than NetApp, EMC, Hitachi, HP, etc? If we would all be honest, they aren't doing it cheaper, but they are spending somebody else's money to get you to buy their product at a subsidized price so they can make a name for themselves. What support structure is behind this startup? A handful of folks? What happens when you have a PB of data and have a 9 alarm fire problem at midnight on a Saturday night that spans storage, OS , and application? Don't call Dell because it is outside of M-F 9-5:-) What about these startups if they go belly up? Who ya gonna call then? Ghostbusters? That will be the time you wish you had NetApp or EMC or Hitachi, or HP and a good relationship with your sales team so you can call and chew them out and get the benefit of a multi-billion dollar organization behind your problem as painful as it may be in this always on world we now live in. NetApp support stinks in some ways and encyclopedic documentation also stinks, but in the end, if you can get the elephant moving, it can solve some very complicated problems. If you can get around all this, you can have some very competent storage doing some very cool things and have a solid support structure behind your investment when you have issues (because Mr. Murphy still lives and thrives despite our insistence that we get 99.999999999 uptime because we are human and fallible no matter what marketing ever says) NetApp support is not perfect and maybe a little too trimmed and offshored for my taste, but still there nevertheless. And since NetApp is a global player, they do need support folks around the globe as not everybody speaks English or has the same mannerisms and customs.
Is NetApp complicated? YES! And for good reason after millions of lines of code from many brilliant minds. Could NetApp make things easier? It sure would be great if they did, but when they do, many of us will be out of a job, just remember that. And if it is made so simple a monkey can run it, what happens when it stops working....will the monkey fix it and will they understand it and all the complex interactions? The transition from 7-mode to cDOT IS painful, but NetApp has released tool improvements (and will continue to do so I trust) to help such as CFT 7MTT and FLI, but is this work just point and click? NO!!! That is why there are PSEs trained and experienced in migrations to help customers If they don't have the skills in-house. cDOT is not that hard to learn and not much different than 7-mode overall at this point, so 7-mode folks, stop being afraid of it, get the simulators, and start working with it and get NetApp to throw in some training credits when you buy your hardware. The best is yet to come if the company can make it and customers see the value. No change of storage platform is going to be pain free, you just choose your pain. If NetApp customers would be honest, some of the pain of migration comes form companies skimping on training and not hiring quality storage professionals and now are reaping what they sowed when it comes time to migrate, because NetApp assumes everybody does things according to those encyclopedic documents....granted some of it is NetApp changing the way some things work, but just like Microsoft, you have to let XP and Windows 2003 die some day to get Windows 10 and Server 2016 (not sure it is "better" but certainly new and needed features exist so hopefully you get the analogy). If NetApp did a better job of taking care of customers and provided services to make sure the customer does things according to best practices through regular customer contact and consultation and some free PSE work (Not free for NetApp but for what NetApp makes, they need to build this in to add value IMO to justify the cost), perhaps customers might not be so "messed up" and difficult to migrate, or fearful of cDOT, and may be easier to keep as customers, but maybe not. I'm not an MBA or IVY leaguer so what do I know?
Give NetApp some credit...NetApp was "done" with 7-mode, but because of customer fearfulness of moving to cDOT due to legitimate concerns over migration difficulty, they decided to support 7-mode till 2020 and I'm sure that decision was not a "free" one to make...perhaps one reason why NetApp let 1200 folks go today? I don't think you will want to wait till 2020.....CFT is really what was needed and now it is here....it is new; therefore it will have bugs like all software and new processes that will take time to be refined as potholes are hit, just like any product, but bottom line, NetApp is getting there.....and there are always those on the bleeding edge that will jump on the grenade to get there first. We need all kinds in IT, don't we?
Customer, why do you think you can get a monkey to manage storage? I see companies skimp on their storage folks and it costs them a lot because the value that NetApp offers is in it's features and functionality built into cDOT (or 7-mode for that matter) and they treat NetApp like dumb storage because they have not been trained and do it like dumb storage folks do it. Believe it or not, if you hire a solid NetApp engineer, you may get somebody that can also do VMware or other virtualization products because of the tight integrations that are in place (at least for NetApp anyway)....they probabally know a fair amount about networks also because NetApp does NAS.....they may aslo dabble in Windows and UNIX/Linux because NetApp can do them all and you have to know something about the hosts to support the storage relationships.....so instead of a monkey, why not try to find a qualified storage engineer that may be really good at tying your whole infrastructure together because NetApp has been the heart of DCs for a few decades now....and chances are they are intelligent enough that they can learn pretty much anything you throw at them for they did learn NetApp. Does NetApp need to evolve, simplify, and improve? YES, and that takes time, investment and training, not going back to DAS and some white box servers. Will hyper-converged work for some folks? ABSOLUTELY. Will it work for all, not likely. We in IT just spent a couple decades moving away from DAS, just like we spent decades moving from mainframe to client/server and what does virtualization and hyper-converged solutions do but go back to a mainframe type architecture. Because it is the "future" it may be pitched as the be all end all solution may not make it so at all or in all cases. Time will tell this too. Bottom line, to be a good architect you need to develop a solution, not because it is the buzzword of the day or solution of the hour, but does it address the requirements that have been presented (or did you even ask)? Does it solve the customer's problems, or is it just a solution you are pitching and hope at the price point you are offering it at gets the customer to bite? That will work one time, but if you don't/can't back up you solution when the do do hits the spinning propeller encased in a wire cage, you've lost that customer forever if you can't stand behind your solution (unless you are personal friends with someone in high places in the US government and then maybe the US government will hand you a bonus at the taxpayer's expense). I've seen NetApp put their money where their mouth is for customers for problems and for 7-mode to cDOT migration solutions, I've seen NetApp make lemonade out of lemons....what more can NetApp do in difficult situations? How about those start ups that are just about out of VC$? Going to go to Best Buy and get your replacement parts and get the Geek Squad to help are ya? Let me know how that works for you in your unmanned COLO;-) Going to have some rocket scientist write all this cool custom software to make your white box solution work as slick as a whistle for some amount cheaper? What happens when your rock star dies or goes to the next highest bidder for their services and their code breaks down or was 1/2 finished?
Training is important for NetApp, Microsoft, VMware, Cisco, etc...the more you know the better you can do your job, but as we all know in IT, nobody want's to pay for it and for what it costs I can't afford to pay for it, so we get what we get. Wish I was a rocket scientist, but I'm not, but I'm not a monkey either:-) Are there new ideas and new ways to do things, YES, and some of them will make sense and take hold. Those that can't or won't change will become relics of the past....I don't think NetApp is a relic quite yet.
I've worked around NetApp for a long time and would consider myself a cheerleader, but I don't wear rose colored glasses or drink the cool-aid. I just think they have some good ideas which can help many customers solve the storage problems they have. NetApp has built up a portfolio of products to meet varying needs, because they recognized that one size does not fit all: just having it in black isn't going to work in this day and age as it did for Mr. Ford. It has taken NetApp a few years to get sales teams to step outside the FAS box. FAS and AFF are great, but it isn't for every need. E-series, StorageGRID, SolidFire, and other offerings are meant to provide solutions designed for other needs not suited for FAS/AFF. NetApp has cloud offerings to work with major cloud players which allows the flexibility of keeping your data in the cloud or on premises based on what makes sense to the customer at the time which could be different tomorrow and they have that flexibility to change on a dime. If the storage world builds standards that allow any storage to work together, then I'm sure NetApp will be on board as they have helped to create open standards to get us to where we are today. So yes, maybe you should take a hard look at NetApp if all you think of them is FAS....this is not your father's NetApp or was that Oldsmobile? Hopefully NetApp doesn't go the way of Olds though, but time will tell that too.
Oh and one last thought. If you think NetApp should do all this for free or on the cheap or give all their intellectual property away, tell that to the 1200 folks (good folks) that just lost their jobs today or however many it was last year. Tell that to SUN microsystems folks that I run into as I do my work that were displaced from the sale to Oracle. Tell that to the EMC folks that will lose their jobs due to the Dell buyout. Open source is great because we are cheap and love "free" and some rocket scientists have time to burn that made their money during the .com boom (before it bust) or can just come up with this stuff so fast because they are gifted that way, but I have a family that I need to ensure has food on the table and a roof over their head, and free just isn't going to cut it, so I'd appreciate if you would consider hiring good/qualified/experienced storage professionals and NetApp storage where it might make sense in your organization (which may be a lot of places if you look at the whole portfolio), and put NetApp to the challenge with a test plan and POC and then make your decision, rather then all the FUD being slung on these sites which doesn't amount to a hill of beans at the end of the day....and think about all these startups that are burning through cash like its Christmas and what happens when Christmas is over?
Please I am wondering if anyone will be able to assist me with the below issue:
Most of the remote site filers, we don’t have CIFS enabled.
So I am wondering if I could get some documentation which talks about without CIFS server setup, or if someone can give me the step by step set up.
Thank you.
Stan
The Register 
Biting the hand that feeds IT
Copyright. All rights reserved © 1998–2024
