Amazon Outage

This message was edited 1 time. Last update was at 2 Jul 2012 00:09:34

I am experiencing some issues after the recent power outage. After the outage I first logged into my PostgreSQL console and noticed it said there were two instances running. However when I clicked on the clusters tab it appeared that I had no clusters running. So I logged into my Amazon console and noticed that there were actually nine instances running. It seems like there was some sort of failure loop that was trying to recover.

Luckily I am in a non-production situation. So I decided to just terminate all the instances (note I also deleted the associated security group). Then I went back to my PostgreSQL console and tried to restore from a backup. However when I do it says that the cluster name I am using already exists. Note that I am using the same cluster name as the old cluster.

So I would just come up with another name except I am concerned that even though no cluster shows, that I am being charged for something that is not operational. Note that the dashboard still shows I have two instances running.

It would seem that something needs to be cleared out which would presumably allow me to restore from a backup to the same cluster name and ensure that I'm not incurring any invalid charges.

Any assistance provided would be greatly appreciated.

Note my cloud database ID is the same as my email address for the forums. Is there a better way to identify myself?

Thanks,
Grant Hiebert

UPDATE: I ended up creating a new cluster from a backup with a different name. My dashboard now correctly shows one instance running so everything appears to be working properly. So false alarm as long as my bill looks good.

The instances are taking longer than usual to actually terminate and clean up since the outage, so the cluster name could not be reused until the clean-up of the terminated cluster and its resources was complete.