On Thu, Feb 17, 2011 at 11:21:26AM +0100, Philipp Hahn wrote:
Hello,
I have a problem with the following code fragment, which refreshes a directory based storage pool:
storage_backend_fs.c#virStorageBackendFileSystemRefresh(...) ... while ((ent = readdir(dir)) != NULL) { ... if ((ret = virStorageBackendProbeTarget(...) < 0) { if (ret == -1) goto cleanup; ... } ... } closedir(dir); ... return 0; cleanup: ... return -1; }
This disables the whole pool, if it contains an Qcow2 volume while, whose backfile is (currently) unavailable (because, for example, the central NFS store is currently unavailable or the permissions don't allow reading). This is very annoying, since it's hard to find the Qcow2 volume, which breaks the pool. I use the following command to find broken volumes: find "$dir" -maxdepth 1 \( -type f -o -type l \) \( -exec kvm-img info {} \; -o -print \) 2>/dev/null Even worse, you can't delete the broken or re-create the missing volume with virsh allone.
To reproduce: dir=$(mktemp -d) virsh pool-create-as tmp dir '' '' '' '' "$dir" virsh vol-create-as --format qcow2 tmp back 1G virsh vol-create-as --format qcow2 --backing-vol-format qcow2 --backing-vol back tmp cow 1G virsh vol-delete --pool tmp back virsh pool-refresh tmp After the last step, the pool will be gone (because it was not persistent). As long as the now broken image stays in the directory, you will not be able to re-create or re-start the pool.
The easiest 'fix' would be to ignore all errors regarding the detection of the backing files file format. This would at least allow users to still create new volumes and list existing volumes.
I agree, we should be more robust in that respect & treat failure to probe the backing store as non-fatal. Daniel -- |: http://berrange.com -o- http://www.flickr.com/photos/dberrange/ :| |: http://libvirt.org -o- http://virt-manager.org :| |: http://autobuild.org -o- http://search.cpan.org/~danberr/ :| |: http://entangle-photo.org -o- http://live.gnome.org/gtk-vnc :|