The instance I run this blog was slightly impacted a few days ago. All of a sudden I could not ssh into the instance and the Apache was running really painfully slow. It did not really work at all. While I was already fantasizing that my really super awesome new web-2.0-youtube-facebook-twitter crossbreed vKaiser.com had gotten some traction and was overloaded by the publicity, I ended up in the AWS site to see the service status. The service status was fine and my hopes were still high. Then the truth hit me, there were others as well in the forums who had similar issues, EBS based image becomes unresponsive and reboot does not help. Can’t either take a spapshot of the EBS volume, but stopping and starting might help. Just have to prepare for the instance to go down very, very slowly.
So, as I could not take a snapshot and was not particulary interested in using a few days old snapshots, I decided to just shut down the instance and give it the time it needs. Eventually, the server went down and I could restart it just fine. Situation back to normal. This incident could of course have been avoided easily by having a backup system ready or even a load balanced setup if I would have the money to run it.
No luck in getting traction.
