This will be the last Weekly View for a couple of weeks as I am off to Malta with Mrs. PL and PL Junior on our yearly family holiday. I’ll be taking notes whilst there to see what Malta has to do with storage, and if you’re really missing the Weekly View that much in my absence remember that you can catch up on past installments here on the blog.
Having been born in the States, you might think that my favourite television programmes growing up were Diff’rent Strokes, The Dukes of Hazard, or Family Ties. Nope. My father inherited an exceedingly dry sense of humour from my grandfather (z”l) which he then instilled in me …growing up my favourite programmes were actually Are You Being Served?, Monty Python, Fawlty Towers, Yes Minister, and Red Dwarf. For someone who was already viewed by his classmates as a ‘bit weird’, I don’t suppose this helped …and the real problem, in the days before satellite television and BBC America was that the only time we could see these programmes was to watch the four hour block each Friday night broadcast by our local PBS [Public Broadcast System] whom had purchased these programmes from the BBC. To say that my mother was less than enamoured with what she saw as a weekly four hour geek humour fest would be an understatement …and we were often made to video the four hour sessions on to VCR tapes to be be viewed on Sunday when mum went out shopping. Equally, the way the PBS purchased the programmes was somewhat erratic …and in the days before Google and the internet …we were often left to our own devices to piece together the correct order for the programme series. Difficult at the best of times, and winding/rewinding VCR tapes was a less than efficient method.
What does this have to do with Data Storage and Protection?
People sometimes ask me when I’m going ‘home’, by which they mean the USA. Truth be told, London is my home and as my family are spread all over the USA …and I haven’t lived there in any capacity for over ten years …I’d be more likely to be making wine in New Zealand [which I hear is much like England in the 1950s] than hunting in vain for Weetabix in the local Piggly Wiggly. For me the invention of SKY+ has made a huge impact on living and working in the UK …I can search from my favourite programmes, I hit a little red button and hey …presto …it’ll record no bother. Hit the little green button and whammo …I have a series link to record every episode. I can even log on remotely to my SKY+ over the internet using Skyplayer to setup a record on my SKY+ box …from anywhere in the world! No bulky VCR tapes …I can record more than one programme at a time …no need to wind/rewind tapes …sheer unadulterated telly bliss.
But here’s something to think about …if you have SKY+, where are your telly programmes being stored? The simple answer is on a hard drive, but what is the data structure? Do you open up SKY+ and see a directory structure that looks like ‘My Documents’ gone wild? Do you have to search for the file for the programme you want to watch and then double click to watch it? Nope. You have an interface that shows you the programmes recorded (and set to record in future), when they were recorded, and so on …and you simply press the big ‘ol PLAY button and you’re away. The ‘secret sauce’ of SKY+ takes care of the rest via indexing and so forth. Do you worry if your SKY+ box is going to ‘crash’? Do you back it up religiously each night? Not likely as the reputation of the SKY+ product is such that most people treat it like the appliance it truly is and get on with more important things in their lives.
A similar ‘revolution’ is happening within data storage and protection. There are many ways to describe this, and I don’t wish to get bogged down in nomenclature, so let’s call this automated data placement. Put simply, the data storage system in question is designed to understand where the data ‘is’ at a block level and then move/promote/demote as business needs warrant without any administrative interaction nor, most importantly, any disruption to the users of said data blocks. ‘Cloud computing’ is predicated on this very idea, however I feel that this is an important solution for our customers in the datacentre as well.
There are several solutions which make great use of automated data placement, and I don’t have the space to list them all here, but I did want to highlight three solutions briefly which I think our customers want and need.
IBM XiV introduced grid storage to the marketplace and is a massively parallel SATA array which allows for scale out without sacrificing performance. How does it do this? Well, the full answer can be a bit complicated …and we’ll soon have some videos up on Browza+ to explain more fully …but in short the IBM solution is to write data to all drives simultaneously. This allows for tier one performance capabilities at a greatly reduced cost, but some customers have asked …how do we know where the data ‘is’ in the array? Well, there are no storage or RAID groups in the array so you cannot locate it in the traditional sense …but the ‘secret sauce’ of IBM XiV allows for logical drive units [LUNs] to be created with an index effectively understanding where all of the blocks are …not unlike SKY+.
But what if I’ve a customer who likes grid storage architecture but wants the comfort factor of having fibre channel and solid state drives in addition to SATA? Enter EMC Vmax, the next iteration of the DMX family. Working in a similar fashion to IBM XiV, EMC allows a customer to also have fibre channel [FC] disk drives, SATA, and solid state disk drives all in the same array. Why? Well, some customers still feel more comfortable knowing they have ‘supercharged’ storage throughput available to them …and are willing to pay for it …even if they never use it. The Vmax will monitor the workload coming into the array and, if the situation warrants it, move the data up from SATA into FC and finally into solid state drives for performance if needs be. Think of it as taking your SKY+ box and sticking it into a Ferrari …just in case you really really really need to watch Fawlty Towers from zero to sixty in 3.4 seconds.
Finally, we have HDS HCP v3.0, which will be HDS upcoming release to their already hugely popular HDS HCAP content management platform. Given that, in a typical customer environment, better than 80% of stored data is unstrucutured …and of that 80%, much of the data will be dormant not having been accessed for a considerable period of time …content management platforms, sometimes known as ‘archive’, can be hugely useful and reduce storage costs by an order of magnitude. We’ll be running an NDA session internally to help explain what HCP v3.0 does specifically, however what has me excited is that this platform sees the emergence of SKY+ for content management. HCP will be able to support object ‘versioning’ so that I can replicate and/or archive …at the block level …only the bits of data which are evolving. In addition, HCP will allow me to have one archive device …working with multiple storage devices! Think of it as one SKY+ box for the whole neighbourhood. You could easily argue that HCP v3.0 is the first content management for the ‘cloud’ as, by utilising the ‘secret sauce’ of automated data placement fully such that it won’t matter where the data ‘lives’ anymore …HCP v3.0 can handle it anywhere.
As always, I hope that you are excited as I am about the solutions available both in the here and now as well as coming down the line which can help save our customers money …please don’t hesitate to contact me if you require any assistance or need more information on the solutions I’ve discussed.