cloudstack-commits mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From GitBox <>
Subject [GitHub] [cloudstack] melnik13 commented on issue #3194: Suspending a VM before snapshot deletion (see PR #3193)
Date Thu, 30 May 2019 17:44:26 GMT
melnik13 commented on issue #3194: Suspending a VM before snapshot deletion (see PR #3193)
   > Here is a link to an article talking about not using qemu-img to take snapshots of
running VMs (there are lots of other posts/pages agreeing).
   > Should we be suspending the VM before taking the snapshot? Is suspending enough to
prevent corruption or is it just "safer"? We are just very gun shy with snapshots on KVM after
experiencing corruption on several VMs...
   You're absolutely right, but ACS do not use qemu-img to create a snapshot of a running
instance. It runs qemu-img only if the instance is stopped, but for the running one it calls
libvirt to perform the job (and libvirt always suspends the VM).
   The problem occurs when ACS is deleting the snapshot from the running instance's volume.
When the deletion procedure is being performed, libvirt doesn't suspend an instance (well,
the instance is being frozen somehow, but `virsh list` still shows it as running).
   As I think, it's not a bug of ACS, as it seems to be a bug libvirt of QEMU/KVM (by the
way, users of Proxmox are facing these issues too), but our workaround (manually suspending
the instance before the snapshot deletion) could help us to outflank the problem.

This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
For queries about this service, please contact Infrastructure at:

With regards,
Apache Git Services

View raw message