Technology changes at an incredible pace, capacity, performance, scale and capability all growing whilst cost per unit drops just as fast. The general operational principles of how to manage and maintain IT infrastructure remain remarkably constant. Sometimes we forget these principles because the cost of ignoring them are so low (take storage and archiving for example) but eventually sheer scale forces us to adopt the common sense processes of our predecessors.
Storage demand is growing faster than storage costs are shrinking and so organisations are looking at how to cost effectively archive their stale data into the lowest cost, simplest environment that they can. Amazon Glacier offers a very attractive solution, being tape based, secure and delivered as a service it really fits with the thinking of the modern CTO. The problem is how do we integrate this service into our data centre operational processes and procedures?
We can adopt the sub-optimal approach of having to request access to an archive (which depends on knowing what is in each archive and where the archive is stored). Alternatively we can do what happened in on the mainframe (where disk was expensive and scarce) and search through file and directory metadata and then automatically request that the file(s) are made available online.
Avere’s new Cloud NAS automatically migrates stale data to Amazon S3 or Glacier based on usage patterns and provides permanent metadata stubs in the Avere unified filesystem to enable automated recovery. Simple, effective and useful all in one solution.
In fact the overall Avere product is much more than an archiving solution, data is identified as being random or serial access and fresh or stale. The Avere filesystem automatically manages placement of the data blocks to optimise performance and cost simultaneously. One unified NAS solution that offers high performance Flash and DRAM based storage for virtualisation and Database type workload, SAS or SATA for sequential reads and writes and cloud storage for very stale data that may never be needed again.
A really neat solution to a perennial problem faced by CTOs around the world.