Replacing Storage? What To Look For

In our last entry, we discussed how the cost of out-year maintenance will cause you to look for another storage vendor and based on the comments and emails, we struck a nerve. Beyond that issue though what else will make you switch storage vendors?What are some of the capabilities you should be looking for in your next storage system? In general there is no single feature that is constant across the board. It will vary from data center to data center. Clearly most sites are looking to drive down costs and improve performance. Most vendors will claim to offer that; you will want to dig deeper.

We typically suggest that the first capability to consider is will this system make you more efficient in managing storage. As we know all too well storage capacity and performance demands are growing everyday. One of the best ways to combat this is to look for systems that are automated. That takes some of the storage management burden off of your back and handle it themselves. This means tasks like provisioning, performance optimization and even drive repair should be able to be handled automatically by the system. In short it should just work. You should no longer have to babysit your storage you should be able to manage it.

Second a new system should not require more capacity than your current system. Several vendors now are advancing thin provisioning to the point that it can also perform a thin migration. Thin provisioning means that when you create a new volume on the storage area it only claims physical space on the drives as data is actually written to them. The problem in the past has been, as we describe in our Thin Provisioning White Paper, this capability has only worked on net new volumes. When you migrated old volumes from an existing array, you typically do a block copy and that means blocks of deleted data and empty space were copied over as well. New storage systems and some modern volume managers have the ability to zero out and identify these areas during the migration. As a result saving capacity during the migration and making fat volumes thin. A few vendors are now providing a guarantee around about this capability. This will reduce your new capacity requirement by X% when you convert to their storage system. Take advantage of those offers.

The final capability to look for is automated tiering, the ability to move data from one storage type to another. For example migrating active data to Solid State Disk (SSD) or inactive data to SATA. Since the early days of Hierarchical Storage Management (HSM) and later Information Lifecycle Management (ILM) we have been trying to figure out how to correctly leverage different types of storage. Now with SSD we have more storage types than ever and of course the quantity of data and number of files to manage have grown significantly. Automated tiering allows the storage system, with override from you, to make the decisions on where to place the data. This is especially critical in SSD, where because of its premium cost, you want to keep it full and only have the most active data on it.

These are just three suggestions of capacities to look for. Don't consider it an all inclusive list nor as a must have list. For many data centers these should be on the list but there can be compelling reasons to live without one of them. Sometimes you don't have the budget for a new storage system. In our next entry we will explore how to get these capabilities by enhancing instead of replacing your current storage system.

Track us on Twitter: http://twitter.com/storageswiss

Subscribe to our RSS feed.

George Crump is lead analyst of Storage Switzerland, an IT analyst firm focused on the storage and virtualization segments. Find Storage Switzerland's disclosure statement here.