How long before containers become the new virtualisation layer?
Imagine a world where you could look at a dashboard and get a complete overview of your data; How much you had, where it was, how much it was costing you, if it was secure, how to optimise it, which were the most valuable elements and all this was done automatically for you! Hopefully this will be something that will come along in the future, but now think about how your data estate currently looks; No real idea how much you have, where it resides, how much it is actually costing you, if it is totally secure (especially with some of it now residing outside your firewalls) and which elements are actually valuable to you.
As the forecasted data deluge doesn’t look like its stopping any time soon, the increasing importance of being able to profile and analyse your data is coming more and more to the fore. Vendors are pushing big data analytics like it’s going out of fashion, and apparently companies should be keeping all data they have ever created, just in case they might need to use it in some undetermined time frame. Unfortunately, most companies’ IT budgets are decreasing rather than increasing, to deal with all this additional data. So, we are now at a stage where how data is managed and stored needs to be looked at in more detail, but where to start?
Add to this the fact that the price of storage is coming down, dedupe ratio’s are increasing and storage is generally more efficient than it was 5 years ago, it is quite easy to miss this ticking timebomb within your network. For most companies, the price of storage currently, is going down faster than the growth of their storage, so they are being tricked into thinking the problem doesn’t exist. There will come a point very soon where data growth will start exceeding the decrease in cost of storage, and at that point they will be left with a huge bill for storage that just increases year on year.
Historically it was a little easier, with the majority of data being held at a centralised location, whether that is one, or a small number of data centres, making the management of that data simpler thing to control. Now with data being created in most if not all sites, and collaboration being one of the key business drivers, we have moved to a more distributed data landscape. The problem with this is that most of the tools currently on the market are setup to profile and manage data in as few locations as possible, as they were designed for a time when the data was centralised.
So where to start? Well, the first place a company needs to start is information – which is ironic when talking about data! Without the relevant and correct information about your data how can you possibly make decisions around what to do with it? There needs to be some way of profiling all of your data to be able to get a picture of how much data you have, where is resides, how its used and how old it is. Once you have that information you can start to build up a picture of how data is being used within your organisation. The first, and easiest place to start, is looking at the older data that hasn’t been accessed for a long time and moving that on to a lower tier of storage (likely object storage, cloud or on-site). The next would be to look at what data collaboration needs you have and see what is the best solution to be able to solve them and lastly is to try and identify which parts of your data is the most valuable/right to be doing some for of data analytics on top.
All of these requirements leads to the fact that a Data Lifecycle Management (DLM) platform is becoming much more of a requirement for all organisations. The ability to scan your data, wherever it may reside and then be able to make policy based decisions around where to place it without any manual intervention becomes increasingly important.