(reproduced with simple local disk, no nfs or ceph)
----- Mail original ----- De: "Alexandre DERUMIER" <[email protected]> À: "Martin Maurer" <[email protected]> Cc: "pve-devel" <[email protected]> Envoyé: Mercredi 6 Août 2014 13:35:16 Objet: Re: [pve-devel] qemu 2.1 released I can reproduce it too, First snapshot works, but second snapshot hang, crash the qemu process Maybe it's a bug in internal-snapshot-async.patch. ----- Mail original ----- De: "Martin Maurer" <[email protected]> À: "Alexandre DERUMIER" <[email protected]>, "Dietmar Maurer" <[email protected]> Cc: "pve-devel" <[email protected]> Envoyé: Mercredi 6 Août 2014 12:57:18 Objet: RE: [pve-devel] qemu 2.1 released > Subject: Re: [pve-devel] qemu 2.1 released > > Just to be sure, it's taking snapshot which fail ? > > or the snapshot is corrupt and we rollback hang ? Creating one snapshots with RAM works, just try to create a second one with RAM after a few seconds fails with: __ Formatting '/var/lib/vz/images/100/vm-100-state-ram5.raw', fmt=raw size=4819255296 TASK ERROR: VM 100 qmp command 'savevm-start' failed - client closed connection __ And the VM is stopped. I see this with local qcow2 storage, but also on ceph. Second issue: Creating a snapshot on NFS (qcow2 on NFS) seems to hangs forever with the snapstate "prepare" and the task log shows "Formatting '/mnt/pve/store-nfs /images/119/vm-119-state-ram1.raw', fmt=raw size=2671771648" Martin _______________________________________________ pve-devel mailing list [email protected] http://pve.proxmox.com/cgi-bin/mailman/listinfo/pve-devel _______________________________________________ pve-devel mailing list [email protected] http://pve.proxmox.com/cgi-bin/mailman/listinfo/pve-devel
