Comment 9 for bug 1746630

Revision history for this message
In , Peter (peter-redhat-bugs) wrote :

Fixed upstream:

commit 71d2c172edb997bae1e883b2e1bafa97d9f953a1
Author: Peter Krempa <email address hidden>
Date: Wed May 18 14:58:25 2016 +0200

    qemu: bulk stats: Don't access possibly blocked storage

    If the stats for a block device can't be acquired from qemu we've
    fallen back to loading them from the file on the disk in libvirt.

    If qemu is not cooperating due to being stuck on an inaccessible NFS
    share we would then attempt to read the files and get stuck too with
    the VM object locked. All other APIs would eventually get stuck waiting
    on the VM lock.

    Avoid this problem by skipping the block stats if the VM is online but
    the monitor did not provide any stats.

    Resolves: https://bugzilla.redhat.com/show_bug.cgi?id=1337073

commit 3aa5d51a9530a8737ca584b393c29297dd9bbc37
Author: Peter Krempa <email address hidden>
Date: Wed May 18 14:40:10 2016 +0200

    qemu: driver: Separate bulk stats worker for block devices

    Extract the fallback path that reloads the stats from disk into a
    separate function.

commit 5d2b0e6f12b4e57d75ed1047ab1c36443b7a54b3
Author: Peter Krempa <email address hidden>
Date: Wed May 18 14:17:07 2016 +0200

    qemu: driver: Remove unnecessary flag in qemuDomainGetStatsBlock

    'abbreviated' was true if 'stats' were NULL