Sage Weil [Fri, 12 Mar 2021 20:21:49 +0000 (15:21 -0500)]
mgr: wait for ~3 beacons on startup if mons are pre-pacific
If we are going active and the mons are pre-pacific, they may have the
bug https://tracker.ceph.com/issues/49778 which prevents our modules
metadata (including options) from being updated (until the next beacon).
Wait a bit (6s by default, 3x the 2s mgr_tick_period) to let this
happen.
This allows us to upgrade from broken pre-pacific mons using cephadm,
which may (if orig cluster is <15.2.5) immediately do a cephadm
migration that relies on the mgr/cephadm/migration_current config
option being present in the mon's mgrmap.
Workaround for https://tracker.ceph.com/issues/49778
Sage Weil [Fri, 12 Mar 2021 20:00:49 +0000 (15:00 -0500)]
mon/MgrMonitor: populate available_modules from promote_standby()
This was done in the beacon path, where there is no active mgr and we
get a new entrant, but not for this case where an existing standby is
promoted to active.
This fixes a problem during upgrade where a new (standby) mgr's modules
have a new module option but it is not reflected immediately (not until
the next beacon).
Fixes: https://tracker.ceph.com/issues/49778 Signed-off-by: Sage Weil <sage@newdream.net>
Xuehan Xu [Wed, 10 Mar 2021 07:31:23 +0000 (15:31 +0800)]
crimson/osd: retrive client_requests' prev_op_id right before "start_op"
ClientRequest::prev_op_id should record its immediate predecessor in the
pipeline. If we capture sequencer's last_issued when creating the client
request, it may not represent that predecessor
Samuel Just [Tue, 9 Mar 2021 02:09:15 +0000 (18:09 -0800)]
crimson/os/seastore: add releasing state for segments pending close
This should fix a bug by which we might start scanning a segment a second
time as it is released and possibly even reused resulting in nonsensical
behavior.
Adam King [Wed, 24 Feb 2021 21:13:01 +0000 (16:13 -0500)]
mgr/cephadm: update caps if necessary when getting keyring
If the caps change from the old version to the new one it causes
issues in the upgrade. This allows the caps to be updated. Currently
only seeing this with iscsi but changing it for other as a precaution
Neha Ojha [Tue, 9 Mar 2021 00:48:58 +0000 (00:48 +0000)]
pybind/mgr/balancer/module.py: assign weight-sets to all buckets before balancing
Add an additional check to make sure that the choose_args section has the same
number of buckets as the crushmap. If not, ensure that
get_compat_weight_set_weights assigns weight-sets to all buckets.
Without this change, if we end up with an orig_ws, which has fewer buckets
than the crushmap, the mgr will crash due a KeyError in do_crush_compat().
Sebastian Wagner [Wed, 10 Mar 2021 12:58:37 +0000 (13:58 +0100)]
Merge pull request #39877 from liewegas/cephadm-rgw-simplification
mgr/cephadm: simplify handling for rgw
Reviewed-by: Juan Miguel Olmo Martínez <jolmomar@redhat.com> Reviewed-by: Sebastian Wagner <sebastian.wagner@suse.com> Reviewed-by: Or Friedmann <ofriedma@redhat.com>
This doesn't actually make sense for FuturzizedStore implementations other
than BlueStore, and then only in a diagnostic capacity. It also doesn't
have any actual users at the moment.
Doesn't actually have a TransactionManager implementation, and
it needs substantial changes to adapt to changes in Onode.
We can resurrect it in the future if we want to.
Kefu Chai [Wed, 10 Mar 2021 03:08:47 +0000 (11:08 +0800)]
doc/_themes: fine tune styling
* fix the "dl" (definition list) display in manpage. before this
change, if an item in dl is composed of strings with different fonts,
it is cluttered into a single string. after this change, they are
separated with proper spacing as before. see the "WORKLOAD" section in
https://docs.ceph.com/en/latest/man/8/ceph-syn/ and
https://docs.ceph.com/en/octopus/man/8/ceph-syn/
* reduce the spacing in "ul" (unordered list) in table. before
this change, we have large spacing at end of a list in table. after
this change, the spacing is reduced. see
https://docs.ceph.com/en/latest/start/hardware-recommendations/#ram
https://docs.ceph.com/en/octopus/start/hardware-recommendations/#ram
* increase the spacing between two "ul"s in table. see
https://docs.ceph.com/en/latest/start/hardware-recommendations/#Processor
https://docs.ceph.com/en/octopus/start/hardware-recommendations/#ram
* refs/pull/39408/head:
test: add test for validating cephfs-mirror daemon service status
pybind/mirroring: interface to fetch mirror daemon status
cephfs-mirror: register mirror daemon with service manager
cephfs-mirror: remove unneeded json_spirit header
Reviewed-by: Patrick Donnelly <pdonnell@redhat.com>
* refs/pull/39687/head:
tools/cephfs-shell: continue file listing even on error
tools/cephfs-shell: refactoring of code related listing
tools/cephfs-shell: fix listing of symbolic links
Reviewed-by: Rishabh Dave <ridave@redhat.com> Reviewed-by: Patrick Donnelly <pdonnell@redhat.com>
Sage Weil [Mon, 8 Mar 2021 14:49:14 +0000 (09:49 -0500)]
python-common: continue to allow RGWSpec(realm=r,zone=z)
This is for backward compatibility: an octopus spec yaml can still be
applied to an existing cluster. Note that it might not work on a new
cluster, since cephadm no longer tries to create the realm or zone if
they don't exist.
Sage Weil [Fri, 5 Mar 2021 21:30:20 +0000 (16:30 -0500)]
qa/tasks/cephadm: drop realm.zone convention for rgw
Note that cephadm.py will no longer do anything with rgw realms and
zones. That means that the setup of rgw roles here is only useful
for the default zone and a non-multisite config.
Sage Weil [Tue, 9 Mar 2021 18:15:20 +0000 (12:15 -0600)]
mgr/cephadm: do not prime service cache on reconfig
Ceph daemon reconfig does not need any daemon state refresh since we don't
do a restart--we just rewrite the ceph.conf. This also avoids priming
our cache with a 'starting' state when the daemon wasn't touched.
Fixes: https://tracker.ceph.com/issues/49675 Signed-off-by: Sage Weil <sage@newdream.net>
This is just so we can load up a stored spec after upgrade. We'll silently
drop it, since we have the service_id, and this was only used to generate
that anyway.
Sage Weil [Fri, 5 Mar 2021 20:52:54 +0000 (15:52 -0500)]
mgr/orchestrator: drop $realm.$zone naming convention
- Let users name the rgw service(s) whatever they like
- Make the rgw_zone and rgw_realm arguments optional, so that they can
omit them and let radosgw start up in the generic single-cluster
configuration (whichk, notably, has no realm).
- Continue to set the rgw_realm and rgw_zone options for the daemon(s),
but only when those values are specified.
- Adjust the CLI accordingly. Drop the subcluster argument, which was
only used to generate a service_id.
- Adjust rook. This is actually a simplification and improved mapping onto
the rook CRD, *except* that for octopus we enforced the realm.zone
naming *and* realm==zone. I doubt a single user actually did this
so it is not be worth dealing with, but we need a special case for
where there is a . in the service name.
projects / ceph.git / log