These configs were used for initialization but it is more appropriate to
require setting these file system attributes via `ceph fs set`. This is similar
to what was already done with max_mds. There are new variables added for `fs
set` where missing.
Signed-off-by: Patrick Donnelly <pdonnell@redhat.com>
It's possible for tell osd.* to race against an osd we stopped but the
cluster doesn't know is down yet. In tha case we'll get ENXIO on that
osd and the command will fail.
In this context, we don't care.
Signed-off-by: Sage Weil <sage@redhat.com>
the buildpackages feature is not part of ceph. and its test was moved to
teuthology in 5a43f8d5. let's remove the leftovers from ceph/ceph. BTW,
we have not run this test suite since then. and its dependency cannot be
fulfilled.
Signed-off-by: Kefu Chai <kchai@redhat.com>
* move Thrasher._set_config() to CephManager, and make it a public
method, and rename it to inject_args(),
* use this method instead of using 'tell ... injectargs ...' directly
Signed-off-by: Kefu Chai <kchai@redhat.com>
* refs/pull/18274/head:
mds: fold mds_revoke_cap_timeout into mds_session_timeout
client: add new delegation testcases
client: add delegation support for cephfs
common: remove data_dir_option from common_preinit and global_pre_init
Reviewed-by: Gregory Farnum <gfarnum@redhat.com>
Reviewed-by: Patrick Donnelly <pdonnell@redhat.com>
Right now, we have two different timeout settings -- one for when the
client is just not responding at all (mds_session_timeout), and one for
when the client is otherwise responding but isn't returning caps in a
timely fashion (mds_cap_revoke_timeout).
The default settings on them are equivalent (60s), but only the
mds_session_timeout is communicated via the mdsmap. The
mds_cap_revoke_timeout is known only to the MDS. Neither timeout results
in anything other than warnings in the current codebase.
There is also a third setting (mds_session_autoclose) that is also
communicated via the MDSmap. Exceeding that value (default of 300s)
could eventually result in the client being blacklisted from the
cluster. The code to implement that doesn't exist yet, however.
The current codebase doesn't do any real sanity checking of these
timeouts, so the potential for admins to get them wrong is rather high.
It's hard to concoct a use-case where we'd want to warn about these
events at different intervals.
Simplify this by just removing the mds_cap_revoke_timeout setting, and
replace its use in the code with the mds_session_timeout. With that, the
client can at least determine when warnings might start showing up in
the MDS' logs.
Signed-off-by: Jeff Layton <jlayton@redhat.com>
/usr/sbin is not in non-login PATH on centos. We already do this for
workunits, see commit 0e53f5f38b ("workunit: include /usr/sbin in the
PATH for all commands").
Signed-off-by: Ilya Dryomov <idryomov@gmail.com>
Some extra coverage of the dashboard, including its standby
redirect mode and the publishing of URIs.
Also invoking the command_spam mode of the selftest module.
Signed-off-by: John Spray <john.spray@redhat.com>
added a qa/rgw_frontend directory for civetweb.yaml and the new
beast.yaml. the rgw suites for multifs and singleton now symlink
rgw_frontend/civetweb.yaml. the multisite, tempest and verify suites
symlink rgw_frontend to test both. this doubles the number of jobs in
those suites
Signed-off-by: Casey Bodley <cbodley@redhat.com>
rbd.xfstests task allows spawning xfstests runs on multiple nodes.
Don't unwind task contexts if one of the runs fails -- let the other
runs finish.
Signed-off-by: Ilya Dryomov <idryomov@gmail.com>
* refs/pull/18192/head:
qa/cephfs: test ec data pool
qa/suites/fs/basic_functional/clusters: more osds
Reviewed-by: Patrick Donnelly <pdonnell@redhat.com>
osd will refused to create new pgs, until its pg number is lower
than the max-pg-per-osd upper bound setting.
Signed-off-by: Kefu Chai <kchai@redhat.com>
mgr: common interface for TSDB modules
Reviewed-by: My Do <mhdo@umich.edu>
Reviewed-by: Jan Fajerski <jfajerski@suse.com>
Reviewed-by: John Spray <john.spray@redhat.com>
... that have empty OSD and MDS caps. Don't add a ',' at the
start of OSD and MDS caps.
Fixes: http://tracker.ceph.com/issues/21501
Signed-off-by: Ramana Raja <rraja@redhat.com>
* refs/remotes/upstream/pull/17697/head:
pybind/ceph_volume_client: add get, put, and delete object interfaces
pybind/ceph_volume_client: remove 'compat_version'
pybind/ceph_volume_client: set the version
Reviewed-by: Patrick Donnelly <pdonnell@redhat.com>
* refs/remotes/upstream/pull/16036/head:
mds: improve cap min/max ratio descriptions
mds: fix whitespace
mds: cap client recall to min caps per client
mds: fix conf types
mds: fix whitespace
doc/cephfs: add client min cache and max cache ratio describe
mds: adding tunable features for caps_per_client
Reviewed-by: Patrick Donnelly <pdonnell@redhat.com>
Reviewed-by: Zheng Yan <zyan@redhat.com>
The module self test commands give us a chance to
catch any other ceph changes that change something
that a module was relying on reading.
Signed-off-by: John Spray <john.spray@redhat.com>
Wrap low-level rados APIs to allow ceph_volume_client to get, put, and
delete objects. The interfaces would allow OpenStack Manila's
cephfs driver to store config data in a shared storage to implement
highly available Manila deployments. Restrict write(put) and
read(get) object sizes to 'osd_max_size' config setting.
Signed-off-by: Ramana Raja <rraja@redhat.com>
bluestore_fsck_on_mount and bluestore_fsck_on_mount_deep are enabled by
default. and bluestore is used as the default store backend. it takes
longer to perform the deep fsck with verbose log. so prolong the
revive_osd()'s timeout from 150 sec to 360 sec.
Fixes: http://tracker.ceph.com/issues/21474
Signed-off-by: Kefu Chai <kchai@redhat.com>
This reverts commit f95798b3ad.
The config_path method wasn't available through inheritance as I thought. Oops.
Signed-off-by: Patrick Donnelly <pdonnell@redhat.com>
Pg state maybe all in active+clean when no recovering going on,
so check it again before timedout.
Fixes: http://tracker.ceph.com/issues/21294
Signed-off-by: huangjun <huangjun@xsky.com>
* refs/remotes/upstream/pull/17694/head:
qa/cephfs: kill mount if it gets evicted by mds
qa/cephfs: fix test_evict_client
Reviewed-by: Patrick Donnelly <pdonnell@redhat.com>
* refs/remotes/upstream/pull/17657/head:
mds: optimize MDCache::rejoin_scour_survivor_replicas()
mds: fix MDSCacheObject::clear_replica_map
mds: support limiting cache by memory
common: refactor of lru
mds: resolve unsigned coercion compiler warning
common: use safer uint64_t for list size
common: add bytes2str pretty print function
mds: check if waiting is allocated before use
mds: go back to compact_map for replicas
mds: use mempool for cache objects
mds: cleanup replica_map access
common: add alloc_ptr smart pointer
common: add warning on base class use of mempool
common: use atomic uin64_t for counter
Reviewed-by: Zheng Yan <zyan@redhat.com>
ceph df accounts for pool size, so there is no need to do it in the test.
Fixes: http://tracker.ceph.com/issues/21381
Signed-off-by: Douglas Fuller <dfuller@redhat.com>
This introduces two config parameters:
mds_cache_memory_limit: Sets the soft maximum of the cache to the given
byte count. (Like mds_cache_size, this doesn't actually limit the maximum
size of the cache. It just dictates the steady-state size.)
mds_cache_reservation: This replaces mds_health_cache_threshold everywhere
except the Beacon heartbeat sent to the mons. The idea here is to specify a
reservation of memory (5% by default) for operations and the MDS tries to
always maintain that reservation. So, the MDS will recall caps from clients
when it begins dipping into its reservation of memory.
mds_cache_size still limits the cache by Inode count but is now by-default 0
(i.e. unlimited). The new preferred way of specifying cache limits is by memory
size. The default is 1GB.
Fixes: http://tracker.ceph.com/issues/20594
Fixes: https://bugzilla.redhat.com/show_bug.cgi?id=1464976
Signed-off-by: Patrick Donnelly <pdonnell@redhat.com>
* refs/remotes/upstream/pull/17679/head:
qa: get asok path from ceph.conf
qa: use config_path property instead of literal
Reviewed-by: John Spray <john.spray@redhat.com>
