Note: "permalinks" may not be as permanent as we would like,
direct links of old sources may well be a few messages off.
On Wednesday 11 May 2011 09:43:41 Felix Frank wrote: > Why complicating? > > When you restore a dd image, you return to that point in time, metadata > included. True, but another thing might be the wish to mount the image. Not sure if this will work OK when the metadata is included. Another idea might be to have the DRBD metadata simply externally on a SDD drive, problem solved. > How about > 1. pause and save the VM > 2. disconnect DRBD (passive host should be Secondary) > 3. resume VM > 4. dd from the backing device of the passive host > 5. connect DRBD -> Quicksync That is a good idea, but in Pacemaker context it needs some more steps: the DRBD resource should be told it should not be concerned with one instance of the clone resource not being there any more (disable monitoring for that device). But it is doable I guess. > Of course, if you ever need to restore one of those images, you should > wipe your old active host including its DRBD metadata so it re-joins the > cluster despite the restored metadata being older. Yes, backup is easy, restore will be more complex. > HTH, > Felix thx Felix, B. -------------- next part -------------- An HTML attachment was scrubbed... URL: <http://lists.linbit.com/pipermail/drbd-user/attachments/20110511/09db0f0d/attachment.htm>