Storagebod's Blog

The Oracle Without Answers

It seems that 2011 is going to be the year of hyperbole and hype; what with EMC's 'Breaking Records' launch and now Oracle's 'Game Changing' launch. At least EMC did break some records at their event but it is doubtful whether Oracle changed any games with their launch of a technology which was previously announced by IBM and Fuji a year ago. 

Yes, Oracle have shipped product first but one suspects that Oracle needed something to announce in the storage space, anything is probably better than the deafening silence around Oracle and storage. And it least to catch some kind of interest, it needed some kind of buzz. However Enterprise Tape capacity increases are never going to set the world alight and nearly always one company's announcement is followed very closely by other Enterprise Tape manufacturer announcing a similar increase.

Oracle seem to be pretty bereft of a real storage strategy at present which is no surprise really as Sun were also fairly bereft of a strategy as well but in this era of integrated stacks is this really a long-term tenable position for Oracle? If they are to really leverage the acquisition of Sun, they need one and they need one fairly soon. There is an obvious solution to Oracle's problem; expensive but obvious. And until Oracle bites the bullet, their announcements will continue to be wind and bluster. 

Bod's Annoyances: Server Virtualisation

Server Virtualisation? Is 'Bod simply being obtuse; yes it is entirely possible but before you write this off as lunacy, I do have my reasons. 

Firstly, server virtualisation is not actually a bad thing; it's been around for many years but the landscape today is actually more than a little concerning and I wonder if it will come to bite us really rather nastily over the next decade. The sheer number of virtual servers being built leads me to wonder about the long term sustainability of the current infrastructures and with the current virtualistion==cloud lazy thinking; I am really rather worried.

So how did we come to such a position? The seeds for this were sown in the early eighties when IBM launched the personal computer; prior to this (and yes I am ignoring the early adventures in personal computing by Apple, Commodore, Radio-shack and the likes), computers were large, complex and extremely expensive beasts managed by the computing department; access was very restricted and not many people had access. But with the personal computing revolution; our relationship with computers changed, indeed we actually started to develop a relationship with computers. 

One person mapped to one PC; that's the very definition of personal computing but as these devices became more powerful, it became pretty obvious that they were capable of so much more but the big, multi-user, multi-tasking operating systems which ran on the large corporate mainframes were not available for these personal computers. So we ended up running larger and more complex applications on these devices but using the operating systems available on the desktop; these were single-user operating systems with limited multi-tasking capabilities. So an application per device paradigm became the norm and even when multi-tasking became available and more common; the paradigm had become firmly embedded. 

And the lack of multi-user capablities really hampered the development of a centralised management and secure environment; so it became even more embedded. Add into that, a growth in CPU power far outstripping our ability to use it for a single application; we ended up in the position that we had rampant physical growth but with ridiculously low utilisation figures.

So in order to use this power and to manage it effectively; we have looked to server virtualisation to fix this problem allowing us to build virtual servers which more effectively use the available CPU or more to the point allow us to run multiple single application servers on a single physical server. It also allows concurrent management of these applications to be more easily achieved. Even the rise of the Intel-based Unix systems has not done little to break the paradigm of one application per server but instead of a physical server, we now talk about virtual servers.

It doesn't have to be this way and we only have to look at the way that virtualisation is used in the mainframe environment; you don't talk about running 1000s of server instances, you are more clever about how you virtualise. You may virtualise an environment but it will still only run a few actual virtual machine instances.You may group related applications into a single machine instance but not have those applications running as separate server-instances.

We could do this in the Intel-space; we should move away from the default position of one application per server instance, we should move to the position where we understand why we do so. There are applications which certainly work better when run in massively distributed and parallel environments but this is not every application. A database server can run more than a single database instance; a web-server can run more than a single web-server.

Okay, it's not server virtualisation itself which annoys me but the unthinking use of it and the traps that we are busy building for ourselves.

Turning desktop operating systems into powerful server systems probably was not the wisest move that the IT industry made but the short-term gains were probably too attractive, can we hope that we don't do the same thing with mobile device operating system? Actually, I am not that hopeful really; history does unfortunately have a habit of repeating itself.

Reading the Ruins

So 2010 is nearly done and I must say that 2010 in storage went pretty much as I expected. 3Par, Isilon and Compellent being gobbled up by bigger companies; no special surprises there. New hardware from HDS, NetApp and IBM; no special surprises there either. And more Cloud-washing from everyone; from Public-Cloud to Private-Cloud, everyone wants a piece of the action. I've taken to calling my test-environment at home a Pico-Cloud. 

So what will 2011 bring us? These are some of my guesses and thoughts.

Even More Cloud?

 I expect EMC to continue painting just about everything they do with Cloud-wash and adding a layer of virtual tarnish to the Cloud. Despite all of this, I think EMC are going to be interesting as always.
It's no secret that we should expect some kind of NetApp killer in the coming together of Clariion and Celerra; with the much maligned Celerra-brand to die and it all to come together under some Unified Storage brand. We should also be expecting very tight integration with VMware; I think EMC still feel very hurt that NetApp have managed to do a better integration job with VMware, whether perception or reality. EMC are going to focus some serious attention on NetApp and believe that technically they can now compete; expect the war of words to escalate.

However, having built a NetApp-like Unified Storage platform; EMC have muddied the water again by buying Isilon and yet again have a standalone NAS product. An internal battle brewing within EMC? If they've learnt the lessons from the DG and Clariion; one would hope not but history does have a habit of repeating itself. At the very least, expect some serious marketing spin to try to differentiate the capabilities. Don't believe it for one minute; if the future of most storage is predominantly file with a ever diminishing proportion of block; Isilon become extremely important very quickly to EMC with the new Unified Storage Platform getting squeezed between Isilon and Symmetrix. Granted not a 2011 problem, more a 2013 issue but something worth thinking about when planning investments.

A side question on Isilon, how quickly do EMC try to move Isilon from its BSD roots to their more preferred internal Linux platform? Now, that's a holy-war brewing!

Huge Portfolio?

HP have far too many storage lines and this has left many of their user-base confused and more importantly, I suspect many of their user-base are feeling reluctant to invest more heavily in an HP strategy. 

David Scott is going to have to be ruthless over the next six months and he will have to move fast to get his charges all pointing in the same direction and singing from the same hymn-book. The future of EVA will need to be clarified very quickly; letting the market decide seems to be pretty rash at this point and is adding to the confusion. I'm expecting a statement of direction from them by the middle of the year allowing customers to put together an investment strategy.

HP's problem is how they scale 3PAR's technology both up and down; I'd expect announcements from them on this. Or are they going to have to look at Lefthand for this.

HP's other problem is that they still don't have a really strong file story; they have stories but most of these have as much substance as a Dan Brown thriller. Could HP acquire again; if there were anyone left to buy, that would not surprise me but I don't see anyone who could fill this hole for them.

Net Apprehension

NetApp refreshed their hardware and released another version of Ontap 8; also announced a virtual appliance version of the filer and by all accounts, they've had a stonking 2010 sales-wise. They continue to out-grow their competitors and continue to gather plaudits; so what will 2011 bring?

Well NetApp aren't a company without problems as much as they like to claim and they appear to have got away with a lot; customers are still giving them a fair amount of slack but...

OnTap 8 needs to get tightened up and there needs to be a single mode; this is certainly NetApp's plan and there is some deserved embarrassment when you bring up the current situation. So expect this to be rectified at sometime in 2011 with OnTap 8 cluster-mode to be fully functionally equivalent to 7-mode with the advantage that it is cluster-able. EMC's acquisition of Isilon makes this all the more important and NetApp cannot afford this to be delayed much longer.

NetApp's teams are tired; I would say that the company is struggling to keep up with it's own growth. I've heard this from people inside and out; teams are stretched too thin and the company is going to have to grow the number of foot soldiers. NetApp are not the only company with this problem but they've always run very lean with no fat; admirable in start-up but NetApp can probably afford to grow some girth.

And finally, NetApp need to acquire someone relatively substantial and not only for the obvious reason but also to prove that they can do so successfully. I don't expect NetApp to stray too far from the storage beat with someone like CommVault in their sights but that will probably bring them into a bidding war with Dell.

Ring the Dell?

It appears that Dell want to be the storage game; getting involved in and loosing a very public auction for 3Par and then ending up the year by purchasing Compellent. If you had asked me at the beginning of 2011; I would have bet very heavily on 3Par going to HP and probably not so heavily on Compellent ending up with Dell but it seems a more logical fit than vice versa.

Dell have three interesting pieces of a storage jigsaw right now; Equalogic and Compellent, solid SME products and Exanet, a clustered NAS product. The latter needs some serious investment to turn it into more than an intellectual curiosity; it's market presence is somewhat less than zero and ever since Dell acquired it, it has been invisible. 

Somehow Dell have to turn all of this into a coherent strategy but Dell have a secret weapon; a couple of years ago, Dell bought a small British storage consultancy called The Networked Storage Company. They have continued to run this as an independent entity; TNWSC generally get involved with enterprise customers when they are looking to issue RFPs etc and you can hear the sighs of pain/irritation/annoyance when a vendor finds out that they are responding to a TNSWC tinged RFP. They also run birds of the feather type get-togethers for storage end-users and the agendas always look interesting even if for various reasons I've not managed to get to one yet. I like to think of them as a more sober #storagebeers. 

Dell by now should have a real insight into the problems that enterprise customers are trying to solve; turning this into a product is going to be more challenging but at least they've been listening.

Its Bazza Magic?

IBM's announcement of the v7000 based on SVC technology was one of the more stand-out storage announcements; not because of the technology, most of the industry had been expecting something like this for some time but because it wasn't strangled by the other interests in IBM. Even long-standing rivals were quietly pleased to see that IBM finally came to their senses and realise that the answer to many of their storage problems had been sitting in Hursley for sometime. 

The DS8000 also appears to the array that refuses to die; in a year dominated by zombie-tinged media, the DS8K is almost the zombie array; every-time someone thinks it's dead, up it rises. Unfortunately, like zombies; it will never be popular and people will keep trying to smash it with shovels and whatever else comes to hand.

Next year tho'; XIV should get its long awaited upgrades, time will tell whether these will be enough to see it become a strategic stand-alone product. Infiniband connectivity between the nodes should finally be with us; this was discussed at the IBM launch many moons ago; a 'flash cache' type enhancement to help XIV cope with more random workloads and give the SATA a boost; clustering to improve scalability and I am wondering if we could see 2.5" drives make an appearance to improve floor-space-to-byte density?

I am also expecting to see significant enhancements to v7000; with both some of the artificial limiters coming off but also Storwize making an appearance with block level compression. Obviously, feature enhancements will make an appearance in SVC as well.

Also expect Storwize to make an appearance in the SONAS product range; expect IBM to launch a smaller version of SONAS to compete with their partner NetApp but also aimed at the smaller Isilon deployments. Finally, expect IBM's storage OEM agreements to become less important to them.

Highly Dependable Storage?

Finally, HDS have refreshed their USP platform with the VSP; feature-rich, enterprise-ready and very capable but is it really where the real growth is? I think this is HDS' real problem; how do they show real relevance to the problems that many storage managers are facing?  

I don't believe that they will but I'm expecting them to push very hard on the Hitachi Unified Compute Platform as a catalyst for IT transformation. I think HDS have a very rich seam to mine here in their traditional high-end market place especially with those customers who are very uncomfortable with putting all of their eggs into Cisco's basket

And just perhaps they will finally buy BlueArc? Perhaps we can convince Dell or HP to put in a bid, just to get things moving? 

And Storagebod?

Well, 2010 was the year when somebody finally allowed me to have storage in my job title again; which was nice as my nascent storage team went from having a few hundred terabytes under management from something closer to three petabytes with an ever steepening growth curve. 

Next year brings the challenges of render-farms, 3D HD editting stations, migrating most of our storage into a new building, building disaster recovery capabilities, multiple upgrades and lots more fun I hope.

Happy New Year!!

Martin

 

Supporting Role?

If you work for a vendor and after reading this you feel that this entry is about your company, then you are right! And if you work for a vendor and after reading this you feel that this entry has nothing to with your company, then you are totally wrong!

I've been working in IT now for more than twenty years; I've done a variety of roles but a big chunk of my career has been running support teams and I've done this both sides of the fence i.e both at an end-user but also at a reseller. I know a lot of people who have worked in support on both sides of the fence and pretty much all of us have coming to the same conclusion; over the past ten years or so, vendor support has got markedly worse! Of course, it may be that we are looking at the past through rose-tinted glasses but I do not believe so. I would also say that this is pretty much across the board; this is not a storage problem but it impacts hardware and software support across the board.

There are a variety of reasons for this decline in quality and ironically, it is some of the services that we as end-users have demanded that may have driven the decline in service. 

The Internet is a big driver in the decline of quality; there are huge amounts of support information and detail on line and we as end-users have demanded that the vendor support databases are put on line; unfortunately, this information is often presented in a manner which it is pretty much impossible to effectively search and you find yourself trawling through screens of unrelated sludge before you find the answer you are looking for. But because all of this information has been provided in a 'self-service' portal, this appears to have been used as an excuse to reduce the number of qualified people that you have in a support centre. 

The Internet has also further enabled the stock holding questions from the support centre and having worked both sides, I know that these are holding questions in general.

1) Are you at the latest level? 

2) Can you send me the logs?

3) Can you send me the configuration files?

Now all of these are valid questions in many circumstances but too often it feels that you are dealing with a robot who is working off a script! This is a Support Centre not a Call Centre; there is a difference! And my (and many other's) responses to the questions are

1) Why? Will it fix my problem? Where in the release notes does it talk about something even vaguely related to my problem? And if I'm in a really sceptical mood, can you send me the piece of the code which fixes the problem. And you realise that this is a live production system and just applying fixes is unacceptable, I will get asked all the questions I've just asked you by the change board. And if it makes it worse, it's my job on the line.

2) Why? Will that log help you diagnose the problem because I've already looked at the log and I can't see anything. Do you have any idea how frustrating it is when you have got a disk off-line and marked failed to be asked for logs from the array?

3) Which files do you want to see and why? 

And often, I've already attached the information to the ticket already; so you asking for them again shows that you have not read the ticket or are just working off the script.

Constantly in conversations with our various hardware engineers from all vendors, we discover reductions in headcount, experience engineers being retired early etc and territory coverage per engineer being increased. We are not buying less hardware, we are buying more and there are more systems out there to go wrong. Now arguably, systems are becoming more reliable but there are more of them. And in the world of storage, we have lots of moving parts; disks spin, tapes spool, robots move and these are all things which wear out and break. Disks get bigger and the potential impact of a disk failure and the resultant rebuild times gets ever larger. 

Talking to people who work in the support centres; it appears to be more important to keep the queue within the targets than solving the customer's problem at point of first contact. There is no longer time to do follow-up calls; for example, calling the customer who had the severity 1 call to ensure that they are happy with the fix. 

This is just the tip of the ice-berg and I could rant on and on about this subject; I was ranting about the decline in service years ago and yet it really is not getting better. For example, I am personally aware of four companies this year who have experience major outages due to problems caused by vendor support; it may be that now I have a fairly high profile in the industry that people tell me their war stories but it seems we are on an upward trend. 

I think it's about time that the vendors started to review what and how they are providing support (fix your websites or at least put a decent search capability on it, it is pretty awful that generally I find myself using generic search engines and the site: directive to search your site); it is also about time that they started treating support centre staff with respect and giving them time to do a great job. 

Apple Abandons Enterprise?

So with the announcement of the withdrawal from sale of the 'Xserve'; have Apple finally abandoned the Enterprise? Actually to abandon something, you would have had to have been serious about it in the first place and Apple have never really been serious about the Enterprise and certainly the Data Centre. Pointing people at the Mac Pro or the Mac Mini is not really a serious suggestion is it? 

Apple have never had and I suspect are never likely to have a serious Data Centre presence in the world of Corporate IT but are we going to see Apple attack this from a different angle. We all know that Steve Jobs loves the ability to control the whole stack from top to bottom and I don't think he is yet ready to completely abandon any ambition that Apple have in the Corporate IT space; indeed, I think that he has been heartened by the user-driven success of the iPad and iPhone in the Corporate space and there is a realisation that he cannot compete with the traditional corporate suppliers on their own ground. So, he's going to try to move it to his ground.

The creation of the OSX app-store is simply the first shot across the bows; the ability to manage the applications on your Mac device in the same way as managing them on your iPad, iPhone (or indeed your Android device) will put a huge amount of pressure in Corporate IT suppliers to provide the same desktop experience to their users. I suspect Microsoft are going to come under an increasing amount of pressure from their corporate customers to provide similar functionality to allow application management by end-users. 

But the current Apple app-store does not really provide corporate customers with the level of control that they require and also, it will be completely unmanageable at scale; for this to be a true corporate play, we are going to need corporate app-stores where a corporate end-user can log into and download/install the apps which their employer have licenses for. However, once you get used to consuming applications in such a model; it is simply a better way of managing things. 

And I wonder if this will be Steve's next play for the Enterprise? Or perhaps Apple are truly about to dump the Enterprise and become a consumer-only company. If he doesn't do it; Google will do it as part of Chrome OS, if that ever sees the light of day as a non-beta product.

I suspect even the faithful of the media-world are getting a little nervous with the current behaviours of Apple; for example no FCP refresh until 2012 in a technology area which is changing incredibly quickly with new demands being made all the time is a little worrying for many of them.

2011 is going to be an interesting year for Apple watchers.

Apple's Debt

Apple owes a huge debt to Windows and the PC industry; in fact, without the PC enthusiasts, Apple would have some really interesting problems. People like to point out that Apple has moved over to commodity hardware; which it has mostly but why has that hardware become commodity? It’s because of the PC. 

Steve Jobs makes a big thing about controlling the whole stack, both hardware and the operating system come from Apple; in this way, Apple get to build devices that simply work.  But the world of the PC massively subsidises his hardware development costs; for example, without the PC and especially the PC games players; where would Apple’s graphics cards come from? And there are many more examples in all of Apple’s hardware. 

It would be nice for the fan-bois to acknowledge this debt; there is no doubt that Apple put together some really nice machines, this entry is being typed on a new MacBook Air which is simply the most gorgeous laptop that I’ve ever owned. Is it revolutionary though? Not really, it’s just a really great instantiation of the laptop. And yes, it’s probably overpriced for what it is but then again, you can say that for many things. 

Perhaps including the integrated stacks some of the vendors are pushing…they need to stick Apple logos on them I guess!!

 

Why, oh why?

After a short hiatus; Barry B is back blogging with a vengeance and I like this entry a lot; there's a lot of truth in it and it does reflect some of my own feelings. And it especially reflects many of my feelings/opinions about vendor blogs; I find vendor blogs a lot more interesting when they are blogging about the 'why' of a product as opposed to the 'what' of a product. 

The best blog entries barely mention a product, you know that there must be a product or at least there is a product coming but that's not really important. It's this discussion of 'why' which should really fuel the debate about the future of IT and data-storage. 

Too much time is spent creating answers to questions which don't exist; time and time again you see kit sitting idle in data-centre because it was bought to answer a question which really wasn't there or at least shouldn't have been asked. 

Here are some of them;

Sometimes, the answer to a question which used to be asked but is no longer. If you dig, this is the 'well, we always do it like that' without realising that actually your customer does not want to do it like that anymore.

Sometimes, the answer to 'what are we going to with the Capex budget which we have left at the end of the financial?'

Sometimes, it's the answer to 'How do I hit my end of quarter/year target?' I discount the crap out of everything and offer a special too-good-to-be-true deal.

Sometimes, it's the answer to question which goes, 'We need to buy something but we've not really spent any time thinking about it...'

And of course, it's the answer to 'What happens if we cancel this project?'

But idle kit should almost never be answer to these questions; if as customer, you don't need the kit pretty much immediately (immediately varies somewhat), you shouldn't by buying it. IT kit depreciates too fast in value to have it sitting idle or even worse, turned on and doing nothing. And even in the case of the canceled project; it is unusual today, that a project requirement is so specific that it's infrastructure cannot be re-deployed.

I like listening to vendors talk about the 'why?' but we as customers should also be talking about the 'why' and why the answers we get from our vendors are right or wrong. 

V(per)PLEXed?

So we have VPLEX and despite some scratching of heads as to what it is; it is really quite simple, 

'storage access is further decoupled from storage physicality'

And this really is nothing especially new; decoupling the storage access from storage physicality has been going on for some time. Servers are getting further and further away from their physical disk. We have been adding abstraction layers to storage access for some time, the big question is whether we need another abstraction layer? 

Actually I think that the additional layer is useful; the ability to present 'real LUNS' from 'storage arrays' as a single 'plexed LUN' and keeping these LUNs in-sync might actually be useful in a number of use-cases. I can see it simplifying and transforming DR for example; I can see it making migration a lot easier and I can see EMC heavily leveraging their VMWare investments. I've said it before and I'll say it again; ever since EMC spun VMWare off, they have acted more in concert than when VMWare were wholly owned. 

Is it useful enough to warrant EMC's claim to have invented a new class of storage device?

I think I'll let the vendor bloggers rip themselves to shreds over that.  

I also think it is interesting that they have at long last decided to pretty whole-heartedly support third party arrays; if anything, this makes it an interesting announcement for EMC. Will they sell any? Well, it's going to be an uncomfortable experience for your run-of-the-mill account manager when faced with a Storage Manager who says 'Well you've just re-invented SVC/USP-V etc...you told me that they were rubbish, so why is yours any good?'

I think the heavy-hitters in EMC are going to be very busy supporting their account-teams.

Why I Don't Ignore NAS?

Stephen's blog entry about NAS strikes so many chords with me that I find it hard to disagree with very much he writes in the entry but I am going to disagree with the central premise/question; you should not ignore NAS if you work in a Storage Team. You may hate it but you should not ignore it.

If you ignore NAS as the storage team, you will find your users going very much their own way and you will find that you have a half a dozen little NAS deployments deployed across your corporate network. This causes headaches for everyone; security, network teams, back-up teams all suffer when this happens. And when it all goes wrong, who is going to get called in to rescue the situation? Well, that would be the storage team and who will still get the blame, that will be the storage team.  

Stephen talks about primary data-centre applications but what does this mean? I think that we all have our own prejudices about what constitutes a primary data-centre application but I am going to argue that this might not be a useful definition any more. If the data has value, it needs to be in a secure, supported environment. 

However, it is worth looking at the problem that NAS is often used to solve; NAS is often seen as the simpler, more flexible and agile solution. In smaller SME environments where you do not have a dedicated team looking after storage and you only have a few systems; this can very much be the case. But as soon as your environment grows and increases in complexity, often interoperability issues start to raise their ugly heads. 

Mostly this tends to be in the CIFS/SMB area in a mixed-environment. If there is one thing that you can currently do to avoid unwarranted complexity in your environment, that is do not deploy SMB in a mixed-environment. And even if some vendors allow you to do so, do not deploy NAS shares using SMB and NFS to share out the same volume; the security models will eventually cause you to cry.

A pure NFS environment generally works well; a pure Microsoft SMB environment generally works well. But they need looking after and the storage team needs to work closely with the server teams; the storage team needs to understand more that pure storage, they need to understand the operating systems that they are talking to. I think this is why often storage teams try to avoid NAS, they really want to focus on spinning rust not what the users want to use it for. 

For the storage team, NAS is not simpler; it is more complex. So is it more flexible and agile?

Flexibility can probably summed up as shared-access to common areas. This is harder to achieve with traditional block storage and can bring you into the world of clustered file systems; these have traditionally been fairly complex and arcane to manage and set-up. They are getting better, GPFS is the one that I am most familiar with and it has come on massively in the terms of useability over the past decade. The only problem with most of the clustered file systems is that they are not available for all operating systems but with the growth of Linux and Windows in the data centre, this is becoming less of an issue.

Agility, it is probably quicker to throw up a NAS share than to present traditional block storage. It can use the same infrastructure as your already in place IP network. You do not need a traditional SAN and you do not need to put in extra cards in your servers. You can just put a NAS device on your network and start sharing. iSCSI, however could also be deployed in the same way, just as quickly and arguable more simply as you do not need to worry about integrating networked file-systems into your existing security domains.

Obviously this is rather fraught with danger but many NAS deployments start this way; this agility leads to a long term fragility in the NAS deployment. A NAS deployment should actually be designed with as much care and attention as a traditional SAN.

It is also possibly fair to say that the NAS vendors have paid more attention to their management UIs than the traditional block storage vendors; they have worked to make it simple to configure and administrate. However, the block vendors are now catching up in the simplicity stakes. 

So if block storage can be less complex, as flexible and as agile as NAS deployments; perhaps we can ignore NAS? But if we had ignored the challenges that NAS had brought to the traditional block model then we probably would not have seen the improvements which have reduced complexity, increased flexibility and agility in the world of block storage. 

And with convergence at the network layer coming in the form of DCB; neither NAS or SAN storage can sit on their laurels. I wouldn't ignore either as technologies in my data centre but I wouldn't ignore RESTful storage or clustered file-systems either. 

Automated Questions

The Storage Buddhist's blog about IBM Easy Tier with SATA and SSD here started me thinking. 

And I quote from his blog

'Easy Tier was left to automatically learn the SPC-1 benchmark and respond (again, automatically)'

Very cool and IBM get some good results from this but I have a few questions and this goes out to all vendors who are doing automated tiering of some sort (I include NetApp's PAM here)

1) How long does the array take to learn the workload and how quickly does it take to achieve an optimal state?

This is very important because it impacts any new application deployment. Do I deploy to the fastest disk and artificially throttle the disk to required SLAs? Users are funny you see and if they discover that application performance is degrading over time, even if it is within the SLAs, they generally complain. Or do I deploy to the slowest tier and let the array tune itself to meet the required SLAs? Now, this might be the worst possible time to impact performance and may impact the user acceptance of the application?

Do I have to come up with some artificial way of simulating the production load and run that for a period to enable the array to tune itself prior to any real users? If so, that has impacts on my ability to respond quickly to my user's demands. And my dynamic data-centre becomes less dynamic than I want.

2) How does the automated tiering impact replication?

In traditional replication technologies; one tries to keep the layout of the local and the remote array in sync. This is challenging enough in a non-dynamic environment and is often a manual task. How do you keep the array layout in sync if the array constantly changing it's layout to reflect the load on it? How can you be sure that your remote recovery array is in an optimal state? Do you want to? I suppose the answer is very much dependant on my first question? How long does it take to reach optimal state? Is it hours? Is it days? What is the delta between optimal and non-optimal performance?

If you have to keep the arrays completely in sync; what is the impact of sending array layout changes to the remote array? How much additional network bandwidth do we require?

3) What is impact on restores both from tape and from snaps/clones?

The backup application has no idea how the underlying physical structure has been changed; it will just sequentially restore the blocks on the LUNs it can see. However the array has no idea if the blocks being restored are the hot blocks it moved onto SSD or the cold blocks it moved onto SATA. It could be a right dog's dinner and obviously it will see a completely different I/O pattern in a restore scenario; you might not want your array self-tuning itself at this point.

I suspect that the answers to these questions could be quite complex but I am certainly interested in how the various vendors mitigate some of the risk that I highlight. Automated Storage Tiering is still in it's infancy; I think we've a lot of lessons to learn. 

Or do I worry too much?