When I was working for a global IT manufacturer, I remember walking into one of our largest UK-based customers for a quarterly planning session with their Global Head of Systems Engineering (this was the first time I had met with them). I remember starting the meeting by having an initial review of our portfolio they had invested in and questioning their general take and how it was working for them.
During the conversation, the customer commented on one of our products, which happened to be a well-known & industry-leading Hierarchical Storage Management (HSM) software product. His comment went along the lines of:
“I dislike HSM products. Unfortunately, they are a necessary evil for us and most other organisations like ours. Do I like [insert name of HSM software we had sold them here], no, but it is less rubbish than the others on the market for our needs, so I’m prepared to put up with it!”
It turns out that the customer’s workflow, like many in their market, used a more costly, performance-orientated infrastructure for when data was being created and processed. Once it was finished, the data still needed to be stored but was seldom accessed and never deleted. They used tape to store their data assets and required the HSM to manage this process.
This meeting has had me thinking and always asking the opinions of customers and industry peers about the notion of HSM as a way of meeting Data Lifecycle Management (DLM) strategies. I can honestly say that throughout my years, I cannot recall anyone being ecstatic about using an HSM tool. Generally speaking, most have always been trying to find a way of removing the need to use one!
Now, I am not saying that all HSM tools are bad; in fact, they provide the way to utilise cheaper forms of storage tiers in an automated fashion using policies that you as a customer set. But, the main objection that I have seen with their use is the ongoing management of the HSM tool itself – it usually requires an individual, as per my customer example above, or a team dedicated to administrating it.
So, is tiering tiring?
Completely eradicating the need for storage tiers would be a more simplistic way of storing data; however, this typically comes at a monetary cost that few are prepared to entertain when looking at the value of the data over its lifecycle. However, as time goes on and solutions evolve, plenty of innovative products on the market include some notion of tiering as part of an integrated product, therefore taking the “tiring” nature of administering tiering away somewhat.
What is clear, though, is that having a good handle on your data strategy and a clear understanding of your data lifecycle management policies is a must. It allows you to continuously review how your data is stored, potentially allowing you to exploit better efficiencies delivered by newer technological developments.
Thinking how to get the most out of your organisation’s data? Our team of experts can help you craft the right data strategy: