Archive for May, 2010

Nothing is designed to deliver miracles, especially data storage.

16/05/2010

I got a serious ticking off from Mrs. PL in bed recently.  Wait, it’s not what you think!

I have been trying to get better about not bringing my MacBook Air to bed over the past few weeks reading storage blogs and industry analysis documentation …besides, Mrs. PL knows all too well the glow of a laptop or Blackberry screen and won’t hesitate to have a full and frank discussion regarding bringing ‘work’ to bed.  ‘But it’s not work, I’m reading an aeroplane blog!’ doesn’t cut ice with herself.

Soooo …there I was, thinking Mrs. PL had drifted off to sleep when I thought I would just flip open my MBA for a few minutes and listen to a little live Air Traffic Control from Sydney before falling asleep.

Now, as a bit of background, don’t go acting all surprised that a) I’m a geek, and b) I’m an aeroplane geek.  Indeed, I have blogged about this several times.  Equally, I’ve come to learn that Mrs. PL attempts to hide my Airliner World magazines from me in a vain attempt to save me from my geek self …or perhaps to shield PL Junior from inheriting too many of the geek traits.  Although I submit that geekery is genetic and hey …the Geeks shall inherit the Earth.

But I digress.  I thought Mrs. PL was asleep and a cheeky few minutes on the laptop listening to ATC through my headphones wouldn’t hurt, would it?

Oh, how wrong I was.

Thinking you’ve plugged in your headphones and verifying that you have …two different things, I’m afraid.

Mrs. PL sat bolt upright in bed when she heard ATC guiding a Qantas flight Q432 from London to Sydney.  At full volume.  Out of my laptop speakers.

As this is a family blog I won’t tell you exactly what Mrs. PL said, but let’s just say that I’m fairly certain that my laptop probably wouldn’t fit there but it wouldn’t be for a lack of trying on her part.

Look, I’m an aeroplane geek ‘tis true …and after Mrs. PL calmed down the next day she agreed I could listen to ATC on my iPod on future before drifting off.

But you don’t have to be an aeroplane geek not to have been impressed with the ‘Miracle on the Hudson’ where Captain Chesley ‘Sully’ Sullenberger successfully crash landed or ‘ditched’ his A320 on the Hudson River and saved all 155 passengers and crew.

What’s this got to do with data storage and protection?

One of the most remarkable things about this event …and there are many remarkable things about the six minutes of this flight  …is that the bird strike which incapacitated the airplane occurred two minutes after takeoff at 15:27:01 and Capt. Sullenberger and his crew ran through three page emergency checklist to attempt to save the aircraft …twice …yet were still able to remain calm and ‘ditch’ the aircraft at 15:31 just four minutes later.

When a plane goes down, there is an immediate public outcry for investigation and possibly regulation to ensure that this never happens again, but where is the same outcry to study what went right when a pilot and his crew …aided by Air Traffic Control and other authorities …save all on board in four minutes following a bird strike which turned an A320 into one of the world’s most expensive gliders?  Yes, there were possibly too many things that went ‘right’ that day to count but a concerted effort has been made to try.

Expressed another way, was the Miracle on the Hudson truly a miracle …or a triumph of technology, process, and training?

I often see the same thing occur with technology generally and storage technologies specifically.

1.  Technology is an enabler, but not a saviour.  Leave the #awesomesauce at home.

The A320 wasn’t going to fly or, more to the point …ditch …itself.  But the design of the plane and the safety controls surely contributed to the successful ditching on the Hudson.

As I discussed in a recent post, it isn’t the ‘awesomesauce’ but, rather, what are we proposing to do with technologies such as thin provisioning, data deduplication, automated storage provisioning, automated storage tiering such as FAST, zero page reclamation that matter.  There is no ‘silver bullet’ technology …none of the aforementioned technologies will ‘save’ you …hence the business benefit should be studied closely and articulated properly.

2. People + Documented and Qualified Processes = Best Use of Technology.

Would the outcome have been different had the A320 on that fateful day not been piloted by Captain Sullenberger …who just so happened to have started his career piloting gliders and also helped develop protocols for airline safety?  Who is to say, although it probably didn’t hurt!  But no matter how good a pilot ‘Sully’ is, he’s also the first to have praised the emergency procedures which were verified and kept very close to hand …sometimes referred to by pilots as ‘kneeboards’ which meant he and the copilot were able to run through procedures quickly without having to ‘remember’ what to do next.

It doesn’t matter how proficient with data storage may be, there is no harm in verifying key procedures and then ensuring they are to hand when performing operations such as data recovery, upgrades to a storage array, and so on.  The array will do what you tell it to [normally!], but to not verify upgrade or data migration procedures …and codify them post verification …well prior to key transitions seems folly at best to me.

3. One size rarely fits all.  Are you sure you’ve got the right equipment for the job?

Captain Sullenberger has praised the safety features of the Airbus A320 in several interviews, but I think it is important to bear in mind that much of the context has been around short ‘commuter’ flights.  Would he select the same aircraft given different conditions, such as coast to coast or transcontinental ‘long haul’ flights?  Perhaps yes, perhaps no …but there are charts and matrices which help pilots and airlines select the most appropriate aircraft for flights, payloads, etc.

When discussing data storage …or indeed just about any technology …one of the first questions we should be asking is ‘what are we going to do with it’?  Now, whilst it is perfectly understandable why customers select the technology on a case by case basis in an RFP style format …what I think would be far more effective would be the creation of a service catalogue as a matrix to help customers make such decisions.  The service catalogue is designed to codify workload characteristics …SLA [Service Level Agreement], IOPS/VOPS [desired throughput], RPO [Recovery Point Objective], RTO [Recovery Time Objective], VTB [Value to Business] …and so on.

4. An ounce of prevention could be worth many more pounds than you could imagine.

Everything went right on the day such that, even after the mechanical failure caused by the bird strike, the plane retained structural integrity and performed as the captain expected.  In short, the last thing he needed in the four minutes when he was rapidly making decisions and executing urgently to land on the Hudson was something to go wrong because proper regular maintenance hadn’t been carried out.

Now, I enjoy talking about scale out data storage devices like IBM XiV and EMC Vmax or data deduplication and data compression just as much as the next storage guy but with customers reaping less than 40% utlisation on average from their storage infrastructures there’s no getting away from the benefits of good storage management.  Yes, it may not be as ‘sexy’ as the wizz bang #awesomesauce being touted by competitors and, in some cases our vendor partners, but there is much to be said about the positive business benefits of proactive storage management.  Chris Evans, a.k.a. The Storage Architect not the ginger bloke on Radio 2, is writing a series of great articles on the Four Pillars of Storage Management which are more than worth reading and consuming.

Right, so …with the Icelandic ash cloud looming large my flight to Edinburgh tomorrow to deliver storage Academy and Masterclasses remains in doubt so I think its off to bed …but not before perhaps a few minutes of ATC radio.

Next week …federation is the new buzzword, but is data storage your internal service provider?  No?  That’s not cricket.

Click here to contact me.

Advertisements

Five reasons why catalogue p*rn may be useful.

03/05/2010

As I started to write this blog entry I noticed that, whilst I knew I hadn’t blogged in a few weeks, the inertia of life has well and truly taken over recently as it has been almost five weeks since my last entry.  Along with my normal day to day responsibilities, the last five weeks has seen an unfortunate death in the family, Mrs. PL embarking on the conversion of our garage into a proper playroom for PL Junior …she is a formidable project manager! …the deployment of our first virtualised datacentre [VDC], trying to enjoy the three days of English summer …you get the point!  I’m just now getting back on top of things, more or less, so normal service will resume from this week and I apologise in advance as this week may see a ‘double’ entry as I try to get back to once a week entries each Friday at 17:00.

As we were clearing out the garage in preparation for the conversion I realised that Mrs. PL and I have a strange habit which could get us into trouble some day.  Hold on, bear with me before you delete this more quickly than usual or I get invited for a brief ‘chat’ with HR!  As we were clearing I was struck by the number of old magazines and catalogues we had amassed in the garage and the habit which led to this detritus.

Mrs. PL and I jokingly refer to our nasty habit as ‘catalogue porn’.

It is a nasty habit, possibly worse than biting my nails or picking my nose at traffic lights, which I was convinced  only I was afflicted with before I met Mrs. PL.  Yet one Sunday morning shortly after we got married, as I was filleting the papers and preparing to throw [or hide] the useless catalogues of rubbish we’ll never need or even know how to use …we didn’t have a garden at the time, but I still enjoyed perusing the odd ‘garden scarifier’ or two …when Mrs. PL looked at me and said ‘Don’t throw those away!  I want to look at those’.

It was in this moment, as I looked at her lovingly, I realised that amongst all that we already shared we also shared a perverse desire to look at catalogue upon catalogue of useless items.  I could come out of the closet, free at last to indulge my secret habit in full view of the world safe in the knowledge I was no longer alone.

But that was quite a few years ago and things have moved on a bit for Mrs. PL and I in the strange world of catalogue porn.  We have our favourites, such as Pedlars …otherwise known as ‘I saw you coming dot com’ in our house …and even have a bit of a game challenging each other to find the most outrageous claims and/or marketing hyperbole.  At present I’m in the lead with ‘Nutrileum’ …what in the name of my giddy aunt is that? what, washing my hair with a substance rumoured to require heavy water and centrifuges to produce? …but Mrs. PL is coming up fast with skin cream ‘pentapeptides’ and their related nonsense.

What’s this got to do with data storage and protection?

Whilst there is the obvious space that our catalogues ‘o crap take up prior to our getting round to throwing them away, the real danger posed by our habit is the very real possibility that …sooner or later …we’re going to buy something from one of these ‘I saw you coming dot com’ merchants.

Why is that a danger?  Well, not to put too fine a point on it, but let’s face facts …the scientific ‘fact’ used to sell this rubbish, when ‘facts’ are even bothered to be used that is, is a bit shady.  I say shady …statistically meaningless would be a more accurate description.  One advert actually states ‘of 84 people polled’ …84 people?!  Ideally a poll should have a sample size of 3,000 to 10,000 to be statistically meaningful …and don’t get me started on randomised distribution!  Suffice it to say, most adverts with their ‘polls’ are about as meaningful as me running in to my local and yelling ‘oi, who likes booze?!’.

But I digress.

We find something similar in technology generally and in storage specifically where I think that sometimes we focus too much on the ‘razzle dazzle’ as opposed to the business benefits.  And yes, I admit I sometimes suffer from this as well!  No comments re my obsession with productivity apps and iPads, thanks, I am all too familiar with my own shortcomings.

1.  Make it statistically meaningful for the customer.

I don’ think that there’s anything wrong with talking about the potential positive effects or business benefits of a technology, so long as you either ‘show your work’ or state that the data may not be available as yet.  For instance, if you’re going to talk about data deduplication and the transformative effects it can have on a customer’s environment by reducing costs …and that you have 522 implementations around the world …don’t you think it would be useful to state ‘we have reduced our customer’s costs by xy% on average’ and back that up with field data?  Yeah, me too.  I’ll continue to challenge our vendor partners on this point and we’re also keen to collect this ROI and TCO data ourselves moving forward.

2. Leave the awesomesauce at home, what’s the benefit?

I fully recognise that I’m a geek and, whilst I often think I’m speaking proper English …albeit slightly accented, admittedly …I may not be speaking in a language easily understood by others.  I’m paid to stay in front of technology and understand not just today’s movements but the technologies of tomorrow most likely to benefit our customers.  But.  There’s always a but, isn’t there?  This must be translated into business benefit for our customers, and I’ve noticed that there are folks in technology who would prefer to focus on the ‘I can move a bit faster than you my competitors’ ‘this one goes to eleven’ style of value description.  Some of our customers call this ‘awesomesauce’.  Sunshine is the best antiseptic, so why not be clear about business benefit?  Geeks like me need to understand the ins and outs of the technology, but we shouldn’t lose sight of aligning technology to business value.

3. One size fits all?

Just as I’m sure there are men in the world who can carry off skin tight spangly leggings …I’m not one of them as you’ll likely have gathered.  Equally, I am not aware of a single ‘one size fits all’ storage technology.  IBM XiV might be appropriate for a customer given certain criteria, where NetApp might be more appropriate for another or EMC vMax for yet another.  This is where true consultancy comes into play and, if I’m honest, we need to work with a customer re the immediate and future challenges in an interactive way that RFPs don’t often satisfy.  Early engagement, in my humble opinion, is key.

4.  The only constant in a customer environment is inertia.

Nothing wrong with inertia, but just because we’ve done a good job of engaging with the customer and articulating how we believe we can help we shouldn’t underestimate the power of how things have been done before or presently.  This is where the Solution Centre in Hatfield, our £10m investment in the technologies we’re discussing, comes into its own.  A Virtualised Datacentre [VDC] might very well be the right answer, but there will surely be the need to test workloads and proposed solutions; it most likely won’t be the technology but, rather, people and processes that will catch you out if you haven’t tested properly.

5. Show me the Nutrileum.

As I’m keen to discuss in future posts, I fundamentally believe that VDCs will help to solve many more customer issues than they may potentially introduce.  At a bare minimum I believe that they can reduce IT costs by as much as 30% – 50% or more.  But here’s the rub …we just don’t have the field referenceable statistics to support these claims as yet.  You can rest assured we will collect these stats as we deploy VDCs for our customers, but in the interim I think it is possible to build a business case based upon field referenceable statistics available from components of a VDC.  We should be able to add £xx saved from   server virtualisation plus £yy saved from grid storage plus £zz saved from network convergence and so on to show a cost benefit over a five year period post VDC deployment, and I see nothing wrong with being honest and up front about where these statistics come from.

If you’ve stayed with me to the end, thanks for reading and I hope that this has been useful as, truth be told …that’s the only reason I blog, hoping that what I’ve learned may be of interest and use to others.

Right, Mrs. PL and I have some catalogues to be getting on with so until next time dear reader.

Click here to contact me.