Browsing through my RSS feeds, I read this post by Chris Mellor – apparently storage is turning into a bit of a horror show. On the one hand I see Chris’ point, storage does seem complex and (as others have described) somewhat cumbersome. However that’s the nature of the beast when you have to retain availability of the company’s assets. There is another way to look at it – all this new technology simply represents an opportunity – the flexibility to do things right.
Storage has never been easy. In my 26+ years in the industry I’ve dealt with storage before the advent of RAID when a disk failure meant the recovery of individual tracks and files of data. Early storage arrays were no better and I’m sure everyone can relate a horror story they experienced in their storage management career. Nowadays, we have extremely reliable systems at such low prices; many of the low end storage arrays on the market today are far more reliable and function rich than devices of 10 or 15 years ago.
But that wasn’t Chris’ point. He was talking about making strategic decisions on product and feature set. As usual, I’d always recommend going back to service requirements. For example, virtual environments allow the migration of advanced features into the hypervisor. In many cases this simplifies previous architectures where failover at the LUN level was a complete pain. If the hypervisor manages it, replication can be removed from the equation. Many storage devices today hide the thin provisioning, compression and other space reduction techniques so the user doesn’t even have to care whether they are enabled or not. In any case, these aren’t strictly requirements but rather side benefits of developing an optimally designed solution. Only legacy architectures make these features a challenge to implement.
I would use a few simple rules when embarking on a storage design.
- Make sure your requirements are clearly defined
- Keep it simple
- Reduce the dependency on vendor specific features
- Choose the right place for data management features and the implications of using them there
- Think about migration off a platform as much as you do about migrating to a new platform
- Make sure you understand the true TCO of a solution
- Don’t believe everything your vendor(s) tell you
- Get a second opinion or independent advice
Now I would say the last one, wouldn’t I? But take my word for it, getting a second opinion, even if it’s from a colleague at another company, information in a blog post or a view from Twitter, will prove on of the best decisions you can make. In the meantime, bring on the new technology!
Comments are always welcome; please indicate if you work for a vendor as it’s only fair. If you have any related links of interest, please feel free to add them as a comment for consideration.
Subscribe to the newsletter! – simply follow this link and enter your basic details (email addresses not shared with any other site).
Copyright (c) 2013 – Brookend Ltd, first published on http://architecting.it, do not reproduce without permission.