MDSAuthCaps: print better error message for perm flag in MDS caps
Reviewed-by: Venky Shankar <vshankar@redhat.com>
Reviewed-by: Xiubo Li <xiubli@redhat.com>
* refs/pull/52892/head:
qa: add test to validate periodic checks by async threads
mgr/volumes: periodically check for async work
Reviewed-by: Kotresh Hiremath Ravishankar <khiremat@redhat.com>
* refs/pull/53070/head:
PendingReleaseNotes: add note about 'm' and 'M' periods
doc/cephfs: add note about monthly scheduling and retention period
qa: update minute period multiplier to m
qa: change minute duration specifier from M to m
mgr/snap_schedule: change minute duration specifier from M to m
Reviewed-by: Dhairya Parmar <dparmar@redhat.com>
Reviewed-by: Neeraj Pratap Singh <neesingh@redhat.com>
Reviewed-by: Venky Shankar <vshankar@redhat.com>
* refs/pull/51891/head:
qa: fix cephfs-top tests for the new fields
cephfs-top: include the missing fields in --dump output
Reviewed-by: Neeraj Pratap Singh <neesingh@redhat.com>
vstart_runner.py removes sudo from the command arguments when "omit_sudo"
is not explicitly set to "False". This leads to 2 error messages in output
of vstart_runner.py which are copied at the end of this commit message.
Setting "omit_sudo=False" for the execution of command should remove both
of these error messages.
This also fixes the potential crash of vstart_runner.py when the FUSE
connections directory is not mounted on the machine and it requires
superuser privileges to mount it.
The 2 error messages copied as it is from the log -
DEBUG:__main__:"sudo" was omitted from the following cmd args before execution and logging using function overriding; check vstart_runner.py for more details.
DEBUG:__main__:> sudo mount -t fusectl /sys/fs/fuse/connections /sys/fs/fuse/connections
mount: /sys/fs/fuse/connections: must be superuser to use mount.
Signed-off-by: Rishabh Dave <ridave@redhat.com>
Check if "/sys/fs/fuse/connections" is already mounted before attempting
to mount it again. Doing so leads to unnecessary error messages in the
vstart_runner.py's output (and most likely in logs for teuthology test
runs too) which confuses the users. See end for the error message.
Also, remove "check_status=False" so that the when the commands crashes
the testing is halted immediately. Initially, this must've been added to
ignore errors generated from re-mounting FUSE connections directory
repeatedly.
The error message -
DEBUG:__main__:> sudo mount -t fusectl /sys/fs/fuse/connections /sys/fs/fuse/connections
mount: /sys/fs/fuse/connections: /sys/fs/fuse/connections already mounted or mount point busy.
Signed-off-by: Rishabh Dave <ridave@redhat.com>
kclient doesn't have CEPHFS_FEATURE_MDS_AUTH_CAPS required to
enforce root_squash. Run root_squash tests only for FUSE client.
Signed-off-by: Ramana Raja <rraja@redhat.com>
A kernel CephFS client with MDS root_squash caps is able to write to a
file as non-root user. However, the data written is lost after clearing
the kernel client cache, or re-mounting the client. This issue is not
observed with a FUSE CephFS client.
Signed-off-by: Ramana Raja <rraja@redhat.com>
Test the 'chown' and 'truncate', which will call the setattr and
'cat' will open the files. Before each testing will open the file
by non-root user and keep it to make sure the Fxw caps are issued,
and then user the 'sudo' do to the tests, which will set the uid/gid
to 0/0.
Fixes: https://tracker.ceph.com/issues/57154
Signed-off-by: Xiubo Li <xiubli@redhat.com>
For "fs authorize" command, AuthMonitor.cc checks if permissions is "r"
or begins with "rw". This check is redundant now.
AuthMonitor::valid_caps() runs MDSAuthCaps.parse() which now runs same
check for the MDS caps, regardless of the command.
Signed-off-by: Rishabh Dave <ridave@redhat.com>
Permissions mentioned in MDS caps flags can either begin with "r" or
"rw", or can be "*" and "all". But it can't start with or be just "w" or
something else. This is confusing for some CephFS users since MON caps
can be just "w".
Command "ceph fs authorize" complains about this to the user. But other
commands (specifically, "ceph auth add", "ceph auth caps",
"ceph auth get-or-create" and "ceph auth get-or-create-key") don't. Make
these commands too print a helpful message, the way "ceph fs authorize"
command does.
Fixes: https://tracker.ceph.com/issues/61666
Signed-off-by: Rishabh Dave <ridave@redhat.com>
Replace call to run_ceph_cmd() by call to get_ceph_cmd_stdout() in
method qa.tasks.cephfs.cephfs_test_case.CephFSTestCase.create_client().
run_ceph_cmd() will not return keyring which is wrong.
get_ceph_cmd_stdout() will return the stdout of "ceph auth add"
command, which is the keyring that is expected to be returned by
CephFSTestCase.create_client().
Fixes: https://tracker.ceph.com/issues/62246
Signed-off-by: Rishabh Dave <ridave@redhat.com>
mds: optionally forbid to use standby for another fs as last resort
Reviewed-by: Ernesto Puerta <epuertat@redhat.com>
Reviewed-by: Rishabh Dave <ridave@redhat.com>
This seems to be causing quota related test failures in the fs
suoite. @lxbsz will push a separate PR addressing the issue in
tracker #56397.
Signed-off-by: Venky Shankar <vshankar@redhat.com>
Although setting "omit_sudo" to "False" will have no impact on ceph-fuse
mount command during testing with teuthology (because teuthology code
does nothing whether omit_sudo is set or unset), let's set it to False
so that there's no chance for any bugs in future.
Signed-off-by: Rishabh Dave <ridave@redhat.com>
* refs/pull/52944/head:
PendingReleaseNotes: add a note for `mds_session_metadata_threshold` mds config
test: add test to verify that a buggy client is blocklisted
mds: add perf counter to track number of sessions evicted due to metadata threshold being exceeded
mds: blocklist clients with "bloated" session metadata
Reviewed-by: Robin H. Johnson <robbat2@orbis-terrarum.net>
Reviewed-by: Dhairya Parmar <dparmar@redhat.com>
Reviewed-by: Patrick Donnelly <pdonnell@redhat.com>
MgrTestCase also needs RunCephCmd. If RunCephCmd is inherited by
CephTestCase, instead of CephFSTestCase, MgrTestCase will automatically
inherit RunCephCmd because it inhertis CephTestCase.
Fixes: https://tracker.ceph.com/issues/62084
Signed-off-by: Rishabh Dave <ridave@redhat.com>
* refs/pull/52676/head:
mds/Server: mark a cap acquisition throttle event in the request
Reviewed-by: Venky Shankar <vshankar@redhat.com>
Reviewed-by: Patrick Donnelly <pdonnell@redhat.com>
Reviewed-by: Kotresh Hiremath Ravishankar <khiremat@redhat.com>
Reviewed-by: Xiubo Li <xiubli@redhat.com>
* refs/pull/52547/head:
qa: add test cases for vanilla ops commands
mds: dump locks when printing mutation ops
common/TrackedOp: support overriding the _dump method
mds: remove op field obsoleted by more usable "reqid"
mds: dump metareq_t instead of full op
mds: add lock type to formatter dump of SimpleLock
mds: mark print methods const
mds: drop MDRequestImpl::msg_lock
mds: lock TrackedOp when dumping
mds: avoid recursive locks dumping state
common/TrackedOp: fix race updating description with proper lock
common/Formatter: add support for dumping null
common/Formatter: refactor generating xml name
Reviewed-by: Xiubo Li <xiubli@redhat.com>
Reviewed-by: Venky Shankar <vshankar@redhat.com>
This relies on the new stdin-killer [1] teuthology helper that allows
interacting with the command's stdin.
[1] https://github.com/ceph/teuthology/pull/1846
Fixes: 8bb77ed9e1
Signed-off-by: Patrick Donnelly <pdonnell@redhat.com>
This change causes the program to exit gracefully when stdin is closed
rather than with a Python exception.
Signed-off-by: Patrick Donnelly <pdonnell@redhat.com>
