[jira] [Created] (CLOUDSTACK-6783) ceph: Volume operations can fail if libvirt lost the storage pool

["Wido den Hollander (JIRA)" <ji...@apache.org>]

Wido den Hollander created CLOUDSTACK-6783:
----------------------------------------------

             Summary: ceph: Volume operations can fail if libvirt lost the storage pool
                 Key: CLOUDSTACK-6783
                 URL: https://issues.apache.org/jira/browse/CLOUDSTACK-6783
             Project: CloudStack
          Issue Type: Bug
      Security Level: Public (Anyone can view this level - this is the default.)
          Components: KVM
    Affects Versions: 4.2.0, Future, 4.3.0, 4.4.0
            Reporter: Wido den Hollander
            Assignee: Wido den Hollander
            Priority: Critical
             Fix For: 4.4.0


When libvirt lost the RBD storage pool, could be due to a libvirt restart, the KVM Agent will re-create the storage pool.

When doing so the createStoragePool() method in LibvirtStoragePoolAdapter doesn't return a properly filled LibvirtStoragePool object.

This can lead to situations where the agent for example tries to delete a volume, but that fails:

"Unprotecting and Removing RBD snapshots of image null/7a3bdc9f-5cd3-4192-a814-18acf561f0d0 prior to removing the image
Failed to delete volume: com.cloud.utils.exception.CloudRuntimeException: com.ceph.rados.RadosException: Could not set configuration option key (-22)"

Notice the "null" in the first line, it's because this information is not properly filled in the returned object.



--
This message was sent by Atlassian JIRA
(v6.2#6252)