Add bool kv_sync_in_progress to indicate thread kv_sync_thread
is running or waiting. Only wake up if thread waiting.
Signed-off-by: Jianpeng Ma <jianpeng.ma@intel.com>
Add bool kv_finalize_in_progress to indicate thread kv_finalize_thread
is running or waiting. Only wake up if waiting.
Signed-off-by: Jianpeng Ma <jianpeng.ma@intel.com>
Changed all messages to 10 that are printed repeatedly in an idle cluster.
This makes debugging Python modules easier as the log
is no longer pullued by mgr core messages.
Signed-off-by: Sebastian Wagner <sebastian.wagner@suse.com>
and s/PYTHON_EXECUTABLE/Python_EXECUTABLE/
it's a regression introduced by 5e2bd7fc4dd6d137cf67336d568da5b0d568efca
Signed-off-by: Kefu Chai <kchai@redhat.com>
PgRecoveryEvent should catch both in and out osds now,
not only out osds. Refresh the corresponding variable
name to better reflecting the function changing.
Signed-off-by: xie xingguo <xie.xingguo@zte.com.cn>
There were times that I could observe the num_bytes_recovered field
of pg_stat_t may decrease, e.g., due to the primary switching
(or similar).
Clamp the pg recovery ratio to 0 too since it might go negative.
Signed-off-by: xie xingguo <xie.xingguo@zte.com.cn>
* refs/pull/29072/head:
qa/suites/upgrade: set pg_autoscale_mode=off on existing pools
mon/MgrMonitor: make pg_autoscaler always on
qa/tasks/ceph.conf.template: osd_pool_default_pg_autoscale_mode = off
Reviewed-by: Sébastien Han <seb@redhat.com>
Reviewed-by: Neha Ojha <nojha@redhat.com>
Reviewed-by: Josh Durgin <jdurgin@redhat.com>
* refs/pull/29085/head:
os/bluestore: add slow op detector for collection listing
os/bluestore: parametrize latency threshold for log_latency funcs..
os/bluestore: cleanup around slow op logging.
Reviewed-by: Sage Weil <sage@redhat.com>
* we've moved to cmake 3.5, so no need to hack for newer cmake
* nss dependency is dropped, no need to install it anymore
* pass clang and clang++ using cmake variable to be more consistent
* pass CMAKE_EXE_LINKER_FLAGS as `/usr/local/opt/llvm/lib` is
not in default library paths.
* mention https://github.com/boostorg/atomic/issues/15
Signed-off-by: Kefu Chai <kchai@redhat.com>
allow `distutils_*()` functions to choose the python according to this
parameter, instead of implicitly using the variable of ${PYTHON_VERSION}
inherited from the caller.
Signed-off-by: Kefu Chai <kchai@redhat.com>
use the ones shiped from the latest cmake. which
* enables us to find the recent python intepreter and development files,
* find intepreter and development in a single `find_pacakge()` command,
simpler this way and less error prone.
and to accomodate this change:
* all `PYTHON${PYTHON_VERSION}_*` variables are renamed to
`Python${PYTHON_VERSION}_*` if we use `find_package(Python2...)` or
`find_package(Python3...)` to find python2 or python3 instead of using
`find_package(Python...)`.
* use "2" explicitly when using python2, as `Python_*` variables are not
defined anymore
* when compiling python support of ceph-mgr, continue using `Python_*`
variables. because we find the python interpreter and development
files using `find_pacakge(Python...)` for ceph-mgr.
Signed-off-by: Kefu Chai <kchai@redhat.com>
Test if an unresponsive MDS client session holding no caps is evicted
directly at session_autoclose without being marked as stale at
session_timeout.
Signed-off-by: Rishabh Dave <ridave@redhat.com>
This enables the use of these handy ref templates outside of Message.h.
This is necessary for a future rework of #26348.
Signed-off-by: Patrick Donnelly <pdonnell@redhat.com>
rgw: add missing close_section in send_versioned_response
Reviewed-by: Matt Benjamin <mbenjamin@redhat.com>
Reviewed-by: Mark Kogan <mkogan@redhat.com>
ll_ref can surge up if application repeatly lookup same dentry
and finally cause overflow.
https://tracker.ceph.com/issues/40775 is an example.
in kernel fuse driver and libfuse, uint64_t is used, update
ceph-fuse side to match.
Fixes: https://tracker.ceph.com/issues/40775
Signed-off-by: Xiaoxi CHEN <xiaoxchen@ebay.com>
