The first patch improves the logging for the user. Currently the return
codes from librados are not written to any logfile, which might leave a
user clueless to why the storage pool is not working.
The second patch sets three timeout options in librados. These are useful for
when the Ceph cluster is down. Librados will then not block for ever, but
return a -ETIMEDOUT so libvirt does not block for ever.
Wido den Hollander (2):
rbd: Include return statusses from librados/librbd in logging
rbd: Set timeout options for librados
src/storage/storage_backend_rbd.c | 144 ++++++++++++++++++++-----------------
1 file changed, 78 insertions(+), 66 deletions(-)
--
1.7.9.5