This will be set automatically in __init__ by looping
through kw.items and creating properties on PVolume
with them.
Signed-off-by: Andrew Schoen <aschoen@redhat.com>
rgw: add time skew check in function parse_v4_auth_header
Reviewed-by: Casey Bodley <cbodley@redhat.com>
Reviewed-by: Matt Benjamin <mbenjami@redhat.com>
Reviewed-by: Adam Emerson <aemerson@redhat.com>
Leaving the pv around doesn't keep us from redeploying on
that device because we also wipefs, but explicitly destroying
it makes the output read better.
Signed-off-by: Andrew Schoen <aschoen@redhat.com>
When we run `pvs -a` it returns all pvs, even ones that
have been recently deleted. Because of this pvs that have just
been zapped with `lvm zap` show up and can cause issues if a device
is zapped more than once.
Signed-off-by: Andrew Schoen <aschoen@redhat.com>
If you use the --destroy flag and are zapping a raw device
or parition then zap will destroy any vgs or lvs it finds on that
device.
Signed-off-by: Andrew Schoen <aschoen@redhat.com>
Clang complains tafter recent changes:
/home/jenkins/workspace/ceph-master/src/osd/OSD.cc:8787:8: error: 'DeleteStart' is a protected member of 'PG'
PG::DeleteStart())));
^
/home/jenkins/workspace/ceph-master/src/osd/PG.h:1870:16: note: declared protected here
TrivialEvent(DeleteStart)
^
/home/jenkins/workspace/ceph-master/src/osd/OSD.cc:9136:6: error: 'DeleteSome' is a protected member of 'PG'
PG::DeleteSome())),
^
/home/jenkins/workspace/ceph-master/src/osd/PG.h:1871:16: note: declared protected here
TrivialEvent(DeleteSome)
^
2 errors generated.
Signed-off-by: Willem Jan Withagen <wjw@digiware.nl>
Move the terminal.success to the methods that have access to the
correct osd_id so that they do not print None for the osd_id
in the log message.
Signed-off-by: Andrew Schoen <aschoen@redhat.com>
It's possible our current map is older, we were destroyed then, but in
newer maps our osd was recreated. This happens when the oldest map after
a recreated osd happens to land on an epoch where the osd was marked
destroyed.
Fix by only exiting if one of the newest maps says we are (still)
destroyed.
Fixes: http://tracker.ceph.com/issues/22673
Signed-off-by: Sage Weil <sage@redhat.com>
When initializing the global context, global_init parses command line
arguments and removes ceph global args from the list. As a result they
are not visible for rbd command line parser and global options were
ignored when passing them to subprocesses like nbd or ggate.
Fix this by keeping a list of ceph global init command line
arguments. To build the list we compare the list of initial command
line args and those that remain after global_init parsing.
After this fix it is possible to pass any ceph global option to a
subprocess.
Signed-off-by: Mykola Golub <mgolub@suse.com>
* refs/pull/19803/head:
mds: get rid of unnecessary if condition
Reviewed-by: Patrick Donnelly <pdonnell@redhat.com>
Reviewed-by: Sage Weil <sage@redhat.com>
* refs/pull/19860/head:
cephfs-journal-tool: add usage help info for header set <field>
mds: add error log info
cephfs-journal-tool: add "set pool_id" option
Reviewed-by: Patrick Donnelly <pdonnell@redhat.com>
If we receive a write within the scrub range, abort the scrub chunk and
shrink the chunk size. If we do this too many times do not preempt and
allow the scrub to complete (to avoid scrub starvation due to client io).
Signed-off-by: Sage Weil <sage@redhat.com>
