if the osd_pgtemp message is dropped before monitor receives it, we need
to resend it. otherwise a pg could be stuck in activating state if the
pg creation was withheld by the max-pg-per-osd on the replica, and then
the replica osd removes some existing pg.
Fixes: http://tracker.ceph.com/issues/23610
Signed-off-by: Kefu Chai <kchai@redhat.com>
cephfs-foo-tool from current master crashes immediately it starts.
Introduced by commit 6972273d53 "global: output usage on -h, --help,
or no args before contacting mons"
Signed-off-by: "Yan, Zheng" <zyan@redhat.com>
Fixes: http://tracker.ceph.com/issues/23624
in python3, the stuff read from wire is represented as "bytes", but
the output of "ceph -w" is supposed to be consumed by human. so decode
it as utf-8.
Signed-off-by: Kefu Chai <kchai@redhat.com>
mgr/dashboard: restcontroller: fix detection of id args in element requests
Reviewed-by: Sebastian Wagner <sebastian.wagner@suse.com>
Reviewed-by: Kefu Chai <kchai@redhat.com>
* refs/pull/21039/head:
qa/workunits/fs: test for cephfs rstats
mds: make rstat.rctime follow inodes' ctime
Reviewed-by: Patrick Donnelly <pdonnell@redhat.com>
* refs/pull/21040/head:
mds: mark damaged if sessions' preallocated inos don't match inotable
Reviewed-by: John Spray <john.spray@redhat.com>
Reviewed-by: Patrick Donnelly <pdonnell@redhat.com>
* refs/pull/21132/head:
client: remove getgroups_cb
client: remove _getgrouplist
client: have init_gids just set alloced_gids to true
client: remove init_groups
fuse: handle errors appropriately when getting group list
Reviewed-by: Jos Collin <jcollin@redhat.com>
Reviewed-by: Patrick Donnelly <pdonnell@redhat.com>
Reviewed-by: Zheng Yan <zyan@redhat.com>
* refs/pull/21167/head:
mds: make stopping mds export strays and subtree at the same time
mds: handle client requests while mds is stopping
mds: check projected linkage before migrating stray dentry
Reviewed-by: Patrick Donnelly <pdonnell@redhat.com>
* refs/pull/21255/head:
mds: automatically create snaprealm for old format quota
client: check if mds create snaprealm when setting quota
mds: remove unused MDCache::snaprealm_create()
Reviewed-by: Patrick Donnelly <pdonnell@redhat.com>
Check ms_inject_socket_failures is valid before we use it.
Fixes: http://tracker.ceph.com/issues/23618
Signed-off-by: Brad Hubbard <bhubbard@redhat.com>
When choosing dirfrags for exporting, consider subtrees imported
from other MDS first. It's because that exporting dirfrags chosen
by MDBalancer::find_export() will create new subtrees.
Signed-off-by: "Yan, Zheng" <zyan@redhat.com>
1. change import_pop_map to multimap because subtrees may have the same
popularity.
2. avoid calculating subtrees' popularity multiple times
Signed-off-by: "Yan, Zheng" <zyan@redhat.com>
There were codes that pass mdr->get_mds_stamp() to {hit_inode,hit_dir}.
The stamp can be older than 'last_decay' in decay counters, which can
cause DecayCounter::decay() to malfunction.
Signed-off-by: "Yan, Zheng" <zyan@redhat.com>
the default timeout is none in that case, there are cases where it can hang forever
due to error cases, since this dumps quite a lot of info the logs grow in GB's, with
default timeout of 1200 we can avoid such huge logs and fail sooner. Any tests needing
higher timeout can pass the required value.
Signed-off-by: Vasu Kulkarni <vasu@redhat.com>
