One of our marketing guys always uses the mantra when discussing Tripwire “You are only days, minutes or seconds from a change that could cause your systems to fail”. Now I’m not easily marketed to, I don’t have great abs in less than 7 minutes a day or any collectable porcelain figurines of the Royal family cluttering my abode, but to me this simple marketing statement makes a lot of sense. Change is the major cause of downtime within an environment and the ability to detect and analyse the change that caused the issue quickly is incredibly important.
But Gavin this is a Virtualization blog; what does this have to do with the virtual world I hear you all cry. Well let’s think about risk for a minute. As people rely more on one system to host multiple machines, the risk of downtime due to a bad change multiplies. VMware has great management tools to reduce this risk within their product range–I’m sure that we’ll discuss some of these in the future–but it’s very difficult for VM vendors to escape the fact that a bad change on the Hypervisor could not just impact one system, but all that are hosted on it. By simply having visibility of what changed on the systems I think that a lot of this risk could be mitigated.
Of course, data integrity is not the only thing that can save your virtual world from blowing up and impacting your real world. What about if the configuration that you are monitoring is already in a sad state of affairs? Like keeping the shine on a broken down car, it has very little point. This, my new bloggy friends, will be the subject of my next post.