At a recent meeting of NetBackup customers I heard some interesting comments on the topic of High Availability (HA) and Disaster Recovery (DR) that I would like to share, along with my thoughts on them.
One customer commented “after the earthquake in Virginia, DR became more important.” It is unfortunate that so often people pay no attention to DR planning until after a disaster occurs. It is too late at that point if you are affected by that disaster. It is much better to be proactive about your DR planning, and it shows you are a good steward of your resources.
Some other customers mentioned that VMware’s abilities are now causing some of their application owners to say vMotion is good enough (for HA). They are now using the traditional HA model with clustering software such as Veritas Cluster Server (often with replication and Global Cluster Option) for their applications that can tolerate little down time. This has ramifications for how you plan your HA and DR strategy. You can now implement service level categories to your application landscape much like you do with your backup infrastructure. If you virtualize most of your applications all of them can benefit from VMware’s capabilities (or your chosen hypervisor). Oftentimes the applications that truly must be up are the ones that don’t virtualize well, so you can leverage technologies such as Veritas Cluster Server, Veritas Volume Replicator, and Global Cluster Option to provide local HA along with the ability to migrate to geographically distant locations if site wide outages occur. You maximize your uptime and minimize your cost.
The most interesting comment I heard was that the people were more important than the IT assets. Without the employees to operate the business, you will be challenged to come back online even if you have your IT assets. I don’t know about you, but I thought to myself “duh!” You can spend all the money in the world on your IT infrastructure to make it HA and DR, but if you have nobody to operate it you are dead in the water. Think about this a minute. How are your employees going to get to your DR site? Are they going to fly there on commercial carriers? Suppose their airport of departure is affected and they can’t leave. How are your employees going to man your DR site? Perhaps a telecommuting strategy can be used so they don’t need to physically be there. I’m not talking about your IT folks either; this is your business people. You can see how complicated this is. DR planning is not just about your IT infrastructure, it is part of your overall corporate Business Continuity Planning – how you will get your business back in operation after a disaster. If you fail at this, your business will likely fail too.
Closing out with the customer comments and a good place to start summarizing is that DR planning is in the top 5 priorities for these customers. For some customers it was number one. As more of your applications are exposed to your external customers (not employees of your business) downtime becomes intolerable. Think about your people, not just your IT infrastructure. Start planning for DR today, not after a disaster hits you.