This is a big logic change throughout MDSMonitor,
and additonally some new commands to handle
per-fs variants of what used to be global things.
Signed-off-by: John Spray <john.spray@redhat.com>
When calculated final average bandwidth exceeds 999MB, it starts
to get printed in scientific notation. Bump up the bandwidth
field precision so it'll remain in floating-point format for
a bit longer.
Signed-off-by: Piotr Dałek <piotr.dalek@ts.fujitsu.com>
FSMap encapsulates a collection of MDSMaps, one
for each filesystem, and a collection of standby
MDS daemons that haven't been assigned to
a filesystem yet.
Signed-off-by: John Spray <john.spray@redhat.com>
Unique identifier for a filesystem within
the ceph cluster. Abbreviated to `fscid`,
deliberately chosen to not conflict
with legacy "fsid" identifiers.
Additionally add mds_role_t type which is
a more general form of rank: ranks are
now only unique within a particular fscid.
Roles are a fscid plus a rank.
Signed-off-by: John Spray <john.spray@redhat.com>
Creating this skeleton to include the new
"ceph fs ..." commands. Needs filling out
with prose about purpose and result of
commands.
Signed-off-by: John Spray <john.spray@redhat.com>
This is only useful if you want to test a
single namespace with no standbys. It's easier
to initially have max_mds=1 and increase it
later than it is to go through the ritual
of decreasing max_mds and then deactivating
a daemon and waiting for it to stop.
Signed-off-by: John Spray <john.spray@redhat.com>
Currently, when you create a pool, until the PGs peer you generate a health
error like
8 pgs are stuck inactive for more than 300 seconds
which is inaccurate and misleading. Instead, set the timestamps to the
creation time so that warnings don't appear until it's clear they're stuck.
Fixes: #14952
Signed-off-by: Sage Weil <sage@redhat.com>
Get uid set in g_ceph_context and pass to civetweb where setuid()/setgid()
will be invoked. This Allows for starting civetweb under privileged port
numbers.
Fixes: #13600
Signed-off-by: Karol Mroz <kmroz@suse.com>
Before invoking global_init(), determine which frontends are configured.
If Civetweb is to be used, ensure it will be responsible for dropping
permissions after radosgw startup.
Signed-off-by: Karol Mroz <kmroz@suse.com>
Allows running the global_pre_init() separately in order to parse config
options prior to invoking global_init().
Signed-off-by: Karol Mroz <kmroz@suse.com>
A few more snapshot fixes, mostly around snapshotted inode/dentry tracking
Reviewed-by: Sage Weil <sage@redhat.com>
Reviewed-by: Greg Farnum <gfarnum@redhat.com>
Original configuration is not working
'enabled_backends' should be changed accordingly
Signed-off-by: RustShen <rustinpeace@163.com>
Update rbd-openstack.rst
journal: re-use common threads between journalers
Conflicts:
src/journal/JournalPlayer.cc
src/librbd/Journal.cc
src/test/rbd_mirror/image_replay.cc
src/tools/rbd_mirror/ImageReplayer.h
src/tools/rbd_mirror/Mirror.cc
(merged interface changes to ImageReplayer, and reduced scope for
change to JournalPlayer due to pr #7884 (wip-14663)).
Reviewed-by: Josh Durgin <jdurgin@redhat.com>
rbd-mirror: minor fix-ups for initial skeleton implementation
Reviewed-by: Jason Dillaman <dillaman@redhat.com>
Reviewed-by: Josh Durgin <jdurgin@redhat.com>
