mon: Improve health status for backfill_toofull and recovery_toofull
Reviewed-by: Joao Eduardo Luis <joao@suse.de>
Reviewed-by: Neha Ojha <nojha@redhat.com>
* refs/pull/28561/head:
vstart_runner: upgrade the check for commands to be run as another user
vstart_runner: split unicode arguments into lists
Reviewed-by: Patrick Donnelly <pdonnell@redhat.com>
rgw: add S3 object lock feature to support object worm
Reviewed-by: Adam C. Emerson <aemerson@redhat.com>
Reviewed-by: Casey Bodley <cbodley@redhat.com>
Rectify the condition that checks if command to be issued as another
user using sudo is issued as a single argument after "-c".
Signed-off-by: Rishabh Dave <ridave@redhat.com>
* refs/pull/28194/head:
test_volume_client: declare only one default for python version
test_volume_client: don't shadow class variable
Reviewed-by: Patrick Donnelly <pdonnell@redhat.com>
cephfs binding already automatically converts paths and some arguments to
bytes.
Fixes: http://tracker.ceph.com/issues/40455
Signed-off-by: Patrick Donnelly <pdonnell@redhat.com>
py3 does not have the unicode built-in. Instead, simply try to do the encoding
and catch failures appropriately. If it quacks like a duck...
(Note that cython apparently cheats and allows the unicode check but this is
simpler.)
Signed-off-by: Patrick Donnelly <pdonnell@redhat.com>
Treat backfull_toofull as a warning condition because it can resolve itself.
Includes test case for PG_BACKFILL_FULL
Includes test case for recovery_toofull / PG_RECOVERY_FULL
Fixes: https://tracker.ceph.com/issues/39555
Signed-off-by: David Zafman <dzafman@redhat.com>
There's little reason to support this new tool with python2.
Fixes: http://tracker.ceph.com/issues/40418
Signed-off-by: Patrick Donnelly <pdonnell@redhat.com>
We just took the curmap ref above; do not call get_osdmap() again.
I think it may explain a weird segv I saw here in ~shared_ptr, although
I'm not quite certain. Regardless, this change is correct and better.
(gdb) bt
#0 raise (sig=sig@entry=11) at ../sysdeps/unix/sysv/linux/raise.c:51
#1 0x00005596e5a98261 in reraise_fatal (signum=11) at ./src/global/signal_handler.cc:326
#2 handle_fatal_signal(int) () at ./src/global/signal_handler.cc:326
#3 <signal handler called>
#4 0x00005596f4fe80e0 in ?? ()
#5 0x00005596e5464068 in std::_Sp_counted_base<(__gnu_cxx::_Lock_policy)2>::_M_release (this=0x5596f4b7cf60) at /usr/include/c++/9/bits/shared_ptr_base.h:148
#6 std::_Sp_counted_base<(__gnu_cxx::_Lock_policy)2>::_M_release (this=0x5596f4b7cf60) at /usr/include/c++/9/bits/shared_ptr_base.h:148
#7 0x00005596e543377f in std::__shared_count<(__gnu_cxx::_Lock_policy)2>::~__shared_count (this=0x7f2b25044e28, __in_chrg=<optimized out>) at /usr/include/c++/9/bits/shared_ptr_base.h:1169
#8 std::__shared_ptr<OSDMap const, (__gnu_cxx::_Lock_policy)2>::~__shared_ptr (this=0x7f2b25044e20, __in_chrg=<optimized out>) at /usr/include/c++/9/bits/shared_ptr_base.h:1169
#9 std::shared_ptr<OSDMap const>::~shared_ptr (this=0x7f2b25044e20, __in_chrg=<optimized out>) at /usr/include/c++/9/bits/shared_ptr.h:103
#10 OSD::handle_osd_ping(MOSDPing*) () at ./src/osd/OSD.cc:4662
Signed-off-by: Sage Weil <sage@redhat.com>
When we are doing cache tiering, we are more sensitive to short PG logs
because the dup op entries are not perfectly promoted from the base to
the cache.
See:
http://tracker.ceph.com/issues/38358http://tracker.ceph.com/issues/24320
This works around the problem by not testing short pg logs in combination
with cache tiering. This works because the short_pg_log.yaml fragment
sets the short log in the [global] section but the cache workloads overload
it (back to a large/default value) in the [osd] section.
Signed-off-by: Sage Weil <sage@redhat.com>
This allows exports to STDOUT to use multiple concurrent operations
and also fixes a potential race condition with concurrent callbacks
and file seeking.
Fixes: http://tracker.ceph.com/issues/40435
Signed-off-by: Jason Dillaman <dillaman@redhat.com>
In zstd d8e215cbee03b038fffe74aebad63b625c42f23c
ZSTD_compress_generic() is renamed to ZSTD_compressStream2().
Signed-off-by: Dan van der Ster <daniel.vanderster@cern.ch>
Increase the per-object append flush bytes to 1MiB to mimic the 16MiB
in-memory writeback cache. Finally, default enable the new journal
writethrough until flush option.
Performance results comparing no journal, pre-change, and post-change:
no journal original modified
4K 40310.84 7223.73 27427.27
32K 21530.37 3256.39 10284.05
256K 3159.67 564.73 1627.16
4M 223.03 48.03 91.91
Fixes: http://tracker.ceph.com/issues/40072
Signed-off-by: Jason Dillaman <dillaman@redhat.com>
When set to true, the journal will not attempt to batch appends until
after it receives the the first flush request from the user.
Signed-off-by: Jason Dillaman <dillaman@redhat.com>
Default to disabling writeback-style append flushes unless overridden
by a call to 'set_append_batch_options'.
Signed-off-by: Jason Dillaman <dillaman@redhat.com>
The original flush_interval/bytes/age batching options no longer
actually do anything. Integrate these better with the max in-flight
AIO handling.
Signed-off-by: Jason Dillaman <dillaman@redhat.com>
