News Flash: Flash fails! it also fails differently!
I think it’s really important to have this in your mind as you consider your Flash strategy, I’ve spoken and blogged many times now about how it’s becoming very clear that the biggest disruption being caused by Flash is in the Tier 1 space, Tier 1 is dead, dead, dead. But there are so many Flash storage vendors out there right now, how do you choose which is the right one for you?
If the workload you are considering for Flash is Tier 1 / Mission Critical then I think you have to consider the criteria that you used when you made your original storage purchase decision. Yes, features and capabilities are important, but so are long term measured and demonstrated reliability and proven world class support, probably even more so than any of the technical capabilities. This is where I’ve been seeing an interesting pattern! Often companies are leaving Flash decisions to the technical teams, nothing necessarily wrong with this as these teams are right to assess the technical capabilities of any new technologies, but here’s the issue and I’ve seen it time and time again, people don’t follow up with the next two most important questions…
1) Is this technology proven? how many have been sold? who else is using it for Tier 1?
2) Can this company truly support me should something go wrong?
Having been with NetApp for 9 years now and having witnessed how customers have really scrutinised our credentials as a Tier 1 support provider, the reality of which is that its probably only been in the last few years that we could really claim that our Global Support and Services operation really is at the level people expect for Tier 1. I therefore struggle to understand how any startup Flash vendor, whose entire company is smaller than our Support team alone, could be considered capable of delivering Tier 1 support.
There are a lot of potentially great Flash products out there right now, our EF550 is a one of them, delivering huge performance, based on a system that’s shipped over 750,000 units, this could and should be used for some Tier 1 workloads.
Our June Launch now brings All Flash configurations of our FAS systems, again based on 100’s of thousands of installed units, engineered over 2 decades, with huge performance, 6 x 9’s of availability and all of the Data Management capabilities that you expect from Data ONTAP.
But here’s the kicker! supported by NetApp! we’ve over 13,000 employees and a Support organisation built over 20 years to develop the support that companies expect to keep them running or to get them running as soon as is humanly possible on the rare occasion a problem might occur.
Tier 1 is dead, dead, dead, All Flash FAS is here and NetApp have the support behind it so that you can really consider it for Tier 1 workloads.
Here’s the video…
Disclosure: I work for X-IO technologies.
If I remember correctly NetApp is now in the 22nd year. Started as a very small company in the NAS market against a big player out there, Auspex. Customers in the 90’s trusted NetApp because they had an excellent technology although the customer service organization was nuts.
And now think about a big player like EMC with five times more employees than NetApp, where the customer service organization has the size of NetApp. Where do you find yourself?
History happens. Start-ups come and go but start-ups are niche player, which are doing an excellent job in their niche. Especially when it comes to flash storage.
Arrogance like “I therefore struggle to understand how any startup Flash vendor, whose entire company is smaller than our Support team alone, could be considered capable of delivering Tier 1 support.” can be sometimes disqualifying. Especially when you see that analysts like Gartner positioning NetApp in the lower midrange for reliability and availability for midrange arrays.
Perhaps it is sometimes better to have a bigger customer service organization.
You have confused arrogance for context, I was not taking an arrogant standpoint but rather one of providing context.
You actually make my exact point very nicely in your response, yes in those early days NetApp had a great product, but the support was probably pretty chaotic, the customers that chose us chose us for engineering apps or web apps which in nearly all cases back then were NOT Tier 1. It has taken us the next 20 or so years to engineer systems, a Global Support organisation and the people and worldwide coverage where I think it is absolutely true to say we have earned the very tough credentials that qualify us for Tier 1 workloads.
If there is any arrogance here then I think it’s making the assumption that as a relatively new company, with relatively new products and a relatively small number of people, that somehow you have earned the right to be classed as a Tier 1 provider, this takes time, significant investment and people to earn and I’m sure that someday you will get there.
I do believe you have interesting technology and people will consider it for a number of workloads, the point of my blog was to highlight the additional challenges that come with providing storage to support Tier 1 applications that people absolutely have to consider in addition.
Disclosure : former NetApp employee, current SolidFire employee
When you look at a critical IT storage purchase, related to support I think the things to be asking are:
– how quickly, even if I have Basic support will you route me to a Tier 3 Engineer who will solve my issue?
– can they help me troubleshoot right across my stack even if the storage isn’t at fault?
– what is your commitment to automated testing and QE in your software?
– how quickly could you get a fully regression tested patch or software enhancement that will help me drive my business forward?
Heritage and employee counts are one way of looking at it for sure. I’ve argued the same in the past. I’ve personally seen the good things NetApp used to be able to do as a customer (2001 – 2003) – I’ve also seen the other side.
Good luck with the AFF launch and see you on the battlefield.
I agree these things are important
The focus also has to be on making it a VERY rare occasion where a support call would ever need to go through to a Tier 3 engineer, we should always try to catch these issues in the QA process. But you know what, software is complex, and the ways that people use it adds even more permutations. This is one of the reasons we just built another Engineering Lab in RTP https://communities.netapp.com/community/netapp-blogs/netapp-360/blog/2014/06/16/netapp-s-new-global-dynamic-engineering-lab-in-research-triangle-park-north-carolina-a-reflection-of-a-customer-first-culture
It’s these kind of investments that we are making to ensure that customers do truly get quality they expect from a Tier 1 vendor.
You’ll recognise me when you see me on the battlefield, I’ll be the one helping the customer 😉
Disclosure : former NetApp employee and customer, current Nimble Storage employee
I don’t normally bite on competition blog posts, I usually sit in the background but you got me on this one Matt !
Matt your post comes across as arrogance. You are confusing size with quality !! I remember back in the day we (NetApp) used to beat up good old EMC for having so many products and the effect it had on the both support and R&D budgets. NetApp are a very long way from being the Fast, Simple, Reliable toaster company that served it so well in the past. That software complexity is an ever increasing circle. A great example is a flash strategy that incorporates Cache in the Host, Cache in the controller, Cache in the Disk, 3 x All Flash array – now which one is right for you – step up Mr Customer and place your bet ?? (Not to mention the support that goes along to develop and productise all these offerings).
I love reading the blogs from the ‘Storage Evangelists’ as whilst we all chug on the marketing kool-aid, I realise the true Storage Rockstar’s are the people who are in support solving customer issues and improving customer experience day after day. Having a product that is designed from the ground up not only for Flash but also Support means you can be so much more efficient and not require an army of support engineers. How do I know this ? There isn’t a week that goes by where a customer tells how how world-class support is. I completely concur with ARob’s comments but I prefer a comment from a recent mutual customer on why he prefers to buy Nimble… ‘Performance, Simplicity and the Support is beyond this world’ – Takes me back to the days of Fast, Simple and Reliable !?
Instead of just adding bodies why not innovate in support as well ? That’s what all these so-called ‘niche’ companies are doing…
Here’s something that intrigues me, startups make some incredible claims many that criticise larger vendors, but these are considered bold, a larger vendor makes this type of claim and its called arrogance?
I am very proud of the support organisation we’ve built, as I mentioned in my previous reply we’ve just built another Engineering Lab with over 3,000 racks available for us to test at truly Enterprise levels, this is in addition to the existing 3000 racks already available in RTP. We have almost 3000 people in our support organisation, supporting over 280,000 systems, of which 30,000 are running ‘Mission Critical’ applications. Autosupport that NetApp introduced right back in those early days has been ensuring that many problems are addressed before they ever have any impact on availability, we now have a Hadoop implementation that’s performing extremely advanced analytics against this to look for potential patterns or issues at a whole new level.
What you learn when you ship in these quantities is learnt because we ship at these quantities, I don’t claim that anyones perfect, but I know that we have a support organisation that take great pride in what they do helping our customers.
If you want to take my response as arrogance then that’s unfortunate and as I said in my previous reply, you are mistaking this for context, when you see what we have built in order to support our customers it’s something that no startup has, that’s not arrogance, that’s simply a fact.
Interesting that NTAP are expanding engineering and product range while closing and reducing the support function as part of the recent RIF (sorry…..business realignment!)
I recently had a conversation with an enterprise user on direct Premium Support who had a performance problem and was shunted around (even directed to YouTube) by L1 NTAP support for a week before it got to L2; only to then have an L2 engineer tell him that the case was new to him therefore not in fact a week old 🙂
In terms of the QA process, it took the local escalation manager more the 24 hours to respond and then the case seemed to disappear into a black hole for two or three days!
In the end the issue was caused by an incorrect configuration, but the user experience is what matters!
Quality not quantity!
I asked one of the Directors for our Support operation about the impact of the recent changes, here’s her response
During the recent changes we didn’t reduce our support functions, we consolidated some of our support team locations to make our operations and knowledge sharing more efficient, and as we did that we moved the headcount so we replaced the people we lost in the centralised location in Amsterdam. In addition we actually made some investments into our support organisations this year and increased our overall headcount
Both her and I would definitely like to follow up on the support case you mention as this is definitely unusual, maybe you could send me the case id (firstname.lastname@example.org) and we can look into it