2
votes

I accidentally took snapshot from volume (100GB) that was attached to running instance (as a ROOT device). The snapshot creation took over one hour! After that I created bigger (400GB) volume from that snapshot and everything went fine. Then I:

  • stopped the instance
  • detached the old 100GB device from my instance
  • attached the new 400GB volume to that instance
  • started the instance

Everything seemed OK in AWS control panel (server running and disk status was fine) but I could not connect to that instance via SSH nor HTTP. SSH error: connection refused. So I changed back to old 100GB disk and instance worked well.

My suggestion was that something went wrong because of taking snapshot from running instance. So I deleted the snapshot and took new snapshot from stopped instance (volume). This time the creation of snapshot took only 5 minutes. Then I just created new volume from that snapshot, attached it to instance and started.. still not working.

So my questions are:

  • Can it be possible that even when I delete the old snapshot the data of that snapshot was partially used when making new snapshot?
  • How can I backup my current volume if every time I take a snapshot it uses old corrupted data

All these things were done via AWS control panel.

THANKS!!

1

1 Answers

1
votes

Can it be possible that even when I delete the old snapshot the data of that snapshot was partially used when making new snapshot?

I don't think so. It could be that when you created the first snapshot there were some resources in contention when trying to create it. Thus it took longer. Amazon is not clear about all the resources that they are using under the hood. In a way is good and bad. Good because it takes away most complexities and bad well because you can't see what the hell is going on.

How can I backup my current volume if every time I take a snapshot it uses old corrupted data

You can either stop your instance before taking the volume snapshot or you can just create an AMI (Image) with making sure that the no reboot option is not checked.

See below:

No reboot