Archive for the ‘Information Lifecycle Management’ Category

Is ILM dead? And what does the Queen have to do with it anyway?

13/08/2009

Firstly, I do hope you and yours are having an enjoyable summer break and relaxing break as I did with Mrs. PL and PL Junior recently in Malta.  As I’m sure with most families, to escape the intensity of the sun during the midday we watched a fair bit of television over our holiday and after watching Doctor Who it got me thinking…television hasn’t always been as visually appealing and interactive as it is today.  Indeed, when television was first broadcast, the presenters sat around a table or stood in front of microphones and simply read the same script they would have read on the radio!  Why is that?  Well, put simply…the producers at that time just didn’t understand the power of television as a story telling medium through a more visual context and just tried to use the new technology much in the same way they had radio.  Needless to say people weren’t over the moon at paying what was a serious amount of money at the time to acquire a technology like television that simply did what their existing radios already did.  It wouldn’t be until producers started producing material that wasn’t available on radio that people started buying televisions en masse…one of those key events was the coronation of Elizabeth II on June 2nd, 1953 which saw many people rushing to buy televisions…or huddle around the set purchased by their family or friends.  Many people I have talked to and met since moving to the UK years ago have told me that they remember watching telly for the first time during the coronation!

What does this have to do with data storage and protection?

Information lifecycle management is dead.  You would be forgiven for thinking this is true given the lack of media coverage ILM receives these days, and customers often ask me just what happened to ILM and is cloud computing the new answer to their data storage challenges.  The long answer to ‘is ILM dead’ can be somewhat complicated as I accept that trade magazines and newspapers need to sell copies to make money and, frankly, cloud computing is the current hot topic that everyone seems to be talking about.  Equally, I believe that cloud computing will play an increasingly important role for both SMBs and enterprise customers alike; for SMBs the promise of cloud computing is the agile and cost effective implementation and support of IT without massive overhead, and for enterprise customers cloud computing could prove a useful and economical adjunct for road warriors as well as a worthwhile strategy for data in their private cloud behind corporate firewalls.  In addition, cloud computing is designed to address the storage architecture and make it easier for users access this infrastructure whilst allowing corporate customers to increase their storage utilisation and reduce costs.  Cloud computing doesn’t really answer what to do with the continued creation of unstructured data and, frankly, why should it when some vendors would be happy to become utility companies in their own right with customers paying for storage as they do electricity today?  Cloud computing is a much deeper topic which deserves it’s own discussion and the short answer for me is no, ILM is by no means dead and is just as valid today as it ever was.  Perhaps even more so given the explosion of the growth of unstructured data coupled with sometimes confusing data compliance and retention regulations.  When I relate this view to customers, their natural response is, ‘So why then haven’t more customers implemented ILM?’

Part of the issue with the relative lack of take up in ILM has been in the marketing of ILM to customers by vendors, resellers, and systems integrators; ILM has often been marketed and discussed as if it were a single shrink wrapped product that customers can buy off the shelf and implement quickly.  This strategy was most likely taken due to the fact that ILM can be a confusing and complicated topic to discuss with customers, but we mustn’t lose sight of the fact that Information Lifecycle Management is actually a series of data storage technologies and techniques designed to automate the movement of corporate data from creation to eventual cremation based upon established business rules and/or service level agreements.  This is an important distinction as this leads to two principal reasons customers haven’t implemented ILM; one, they haven’t aligned their corporate data to business value and two, they rightly view the implementation of ILM as potentially disruptive to their production business.  What to do?

The first step, in my opinion, is we need to help customers to align their data to business value.  This is an easy thing to say but, in actuality, becomes somewhat more challenging when you take on board the fact that each customer will be at a different stage of aligning data to business value as well as the fact that each customer often interprets data compliance and retention regulations slightly different.  This is not to say that aligning data to business value is not a worthwhile exercise as I fundamentally believe that this is the cornerstone to the adoption and implementation of ILM.  As with ILM, there isn’t a shrink wrapped product available off the shelf which helps a customer to automate the alignment of data to business value.  The most assured way I have found in helping customers to achieve this alignment is through fixed cost and fixed deliverable consultancy.  These engagements make use of software discovery tools which identify structured and unstructured data, including the hugely useful classification of data, and result in a consultant delivering a report which details a coherent and practical approach to the customer’s implementation of ILM over a set period of time.  Customers can sometimes react cynically to this approach, and my experience is that vendor supplied storage assessments and consultancy have too often given a very limited, immediate, and short sighted approach to recommendations which simply point to the acquisition of more kit.  We must actively work with our customers to help demystify the consultancy process by showing them how we conduct such engagements, how long it will take to deliver, and what business benefits they will receive by investing in such an engagement.  Computacenter have a well defined storage strategy engagement which does just this, and the documentation can be found here.

The second step, following the successful consultancy engagement which establishes how a customer will align their data to business value, is the implementation of technologies which will move data to the appropriate ILM tier in a non-disruptive way.  This is critical as no matter how great an ILM implementation strategy is, and irrespective of the return on investment and cost benefit analysis which supports the strategy, no customer will implement ILM if it is going to be disruptive to their production business.  To be fair, they need to be conducting business and increasing their revenue whilst implementing ILM in a way which is transparent to their internal users and external customers alike.  One of the foremost storage technologies for achieving this movement of data in a non-disruptive way is storage virtualisation.

Storage virtualisation has actually been in the market for longer than server virtualisation, however many customers have not seen the value and import in the implementation of data block level storage virtualisation.  There are several ways to achieve storage virtualisation, and I don’t wish to digress into an esoteric argument about the merits of array versus fabric storage virtualisation.  That’s what the Computacenter Storage Academy bootcamp sessions and related Masterclasses are for!  What I will say is that I believe it is important to consider such technologies which have a proven track record.  There are several Computacenter business partner products such as HDS USPV, the HP XP array which is based on the HDS USPV range, and IBM SVC all have referenceability worldwide as storage virtualisation products capable of virtualising data in a mult-vendor environment.  Again, there are are several applicable options and solutions within storage virtualisation and the key takeaway is the need to move data from tier to tier within existing corporate environments in a non-disruptive way.

I hope that I have been able to prove that ILM is by no means dead…and that we’re not looking to use technologies the same way we have before ala early television adoption…but, rather, ILM has not been taken up by customers in a significant way due to confusing market messages and a perception that ILM implementation will disrupt their business.  I don’t claim to have all the answers, but within Computacenter we are seeing a much great customer acceptance of ILM when we structure the conversation around the alignment of business data to business value as the ‘what’ in the ILM equation and the virtualisation of data to introduce non-disruptive data movements as the ‘how’.  Perhaps you will find similar positive responses through such techniques, but feel free to contact me if I can be of assistance in helping to articulate these messages …and long live Information Lifecycle Management!

-Matthew

Click here to contact me

Have I mentioned I don’t like moving?!

11/08/2009

I have a confession to make.  I hate moving house.  I don’t mean gently despise, gosh I could really do without having to move…we’re talking would rather poke my eyes out with red hot knitting needles hatred here.  You would think that having had to move from my family home to university, from uni to many and several cities in the USA where I worked, from Atlanta to Dublin, from Dublin to London, from flat in London to nice house in Mill Hill, that I would be old hat at this and moving would be no big deal.  Errr, no…couldn’t be further from the truth.  Moving house ranks up there as something that I would do just about anything to have done for me, although even having the work physically done for me if but a small fraction of why I hate moving.  I’ll try to explain why.

I like reading.  A lot.  And so it takes next to no time at all for whereever I might be living to become inundated with books, magazines, newspapers…you get the picture.  But I also have a dirty little secret….I don’t throw away anywhere near as much read stuff as I should!  I read an article/book/whatever and I either know someone that would also enjoy reading this or naievely believe that I will need this article/book/whatever at a later date.  Yes, I know with the internet I could probably find it again in an archive somewhere but logic has precious little to do with this!  I also love music and have a collection of LPs, cassette tapes, CDs, et al that I will probably never listen to again now that I have an iPod.  But have I thrown the old stuff away?  I think you can guess that I haven’t.  And let’s not talk about my obsessive collection of DVDs as I think you are getting the picture.

So what happens when I have to move?  There’s the rub.  If I was going to stay in one place, I would probably just get a skip every few years and cull my collections of reading material, music, and DVDs to make space for more reading material, music, and DVDs.  But I could do this in my own time and at my own leisure.  Moving accelerates this whole process and I then have to stand there with boxes deciding whether to take it with me, cull it, give it away, and so on.  Just thinking about having to do this is enough to make my teeth itch and desperately want a nice cold G&T.

What does this have to do with data storage and protection?

Information Lifecycle Management, or ILM for short, has been one of the biggest buzzwords in storage over the past 6 or 7 years.  What began with storage virtualisation and other similar products designed to increase storage utilisation has resulted in end-to-end ILM vendor portfolios.  I should mention at this point that some folks call ILM Hierarchical Storage Management or HSM…horses for courses, ILM and HSM are pretty well the same thing and I view this in somewhat the same vein as football versus soccer.  It’s football, by the way, but if my wayward American brethren insist on calling it soccer well…let’s just hope they can at least understand the offside rule.  But I digress.  ILM at it’s best is the ability to move data from high tier to mid tier to low tier to archive seamlessly with no manual intervention and without disruption to the user.  Great.  What the heck does that mean?

What if I could have someone [let’s call him Bob] at my beck and call 24/7/365 to manage my books/newspapers/magazines/music et al.  I would tell Bob that, when I’m done with a newspaper or magazine, go ahead and digitise it and catalog it for me on a hard drive.  After I’ve read a book, wait three months and if I haven’t given it to someone go ahead and move it to an off-site storage box but catalog that too so that I know where my books are.  And if I don’t ask about my book for a further nine months after that, digitise it and put it on the same hard drive as my newspapers and magazines.  With the hard drive, make sure that any data that I haven’t accessed in six months gets moved to a lower cost data archive and if I haven’t accessed it from the archive in another six months back it up to a tape and send it to the book storage area.  But remember, Bob…I want to access the data I just mentioned any time I want and I don’t want to wait more than 10 minutes to get it.  Forever.  And all of this needs to happen without my seeing or knowing it is happening.  And don’t eat too much either, Bob…Mrs. PL takes a dim view of that and may make me move you [and me!] out to the garden shed.

Imagine the space I would save!  I would never have to cull again!  I could move at will, or at the very least be able to talk to Mrs. PL about nice houses without dreading the book/music/DVDs to move.  Sheer unadulterated bliss!  Put another way, ILM on a very personal level.

Let’s look at a vendor ILM solution…when EMC began to acquire companies such as Data General, Legato, Documentum, Kashya, and RSA many people were scratching their heads.  Why would EMC, known for their high availability / high throughput systems in the upper end of the market, want to acquire a mid tier disk company [Data General made the CLARiiON platform] a backup software company who also had archiving capability [Legato Networker] a content and document management company [Documentum] and so on?  The answer was they were [and still are] building an ILM portfolio so that they can offer an enterprise solution to a customer which will move data seamlessly, preferably at the block level, between tiers transparent to the users…data storage at the appropriate price point from cradle to grave.  Nirvana!  Does it work?  Well, yes and no.  EMC, IBM, HDS have all been working on ILM and we are a lot further along now than we were five years ago but it isn’t as ‘seamless’ and transparent to the user as it could be.  Thankfully within the CC Storage Practice we understand the ILM portfolios of each vendor and so can implement the solution in a non-disruptive way, and the ILM solutions get better each day.

I’ll leave you with a very interesting ILM solution which has recently entered the market.  Compellent.  Compellent offer ILM ‘in a box’ whereby they have a rack of disks capable of moving data from tier to tier based upon business rules at the block level.  Confused?  Think ILM as described above, but if I only partially read a book with the intention of coming back to it later Compellent would be able to move the pages I’ve already read whilst I would still think that there is an entire book sat on my bedside table.  Clever stuff, and ILM at the block level may very well be the answer in future.

Computacenter services and solutions within Storage and related software are very much underpinned by best of breed vendor ILM solutions, please contact me if we can help you discuss Computacenter ILM solutions with your environment.

-Matthew

Click here to contact me