Kefu Chai [Mon, 8 Feb 2021 13:43:09 +0000 (21:43 +0800)]
crimson/common: mark ExitBarrier final
to silence warnings like:
crimson/common/operation.h:394:20: warning: class with destructor marked
'final' cannot be inherited from [-Wfinal-dtor-non-final-class]
~ExitBarrier() final {
^
Merge pull request #39259 from sebastian-philipp/cephadm-execnet-chmod-600
mgr/cephadm: on ssh connection error, advice chmod 0600
Reviewed-by: Dimitri Savineau <dsavinea@redhat.com> Reviewed-by: Juan Miguel Olmo Martínez <jolmomar@redhat.com> Reviewed-by: Michael Fritch <mfritch@suse.com>
Sage Weil [Thu, 4 Feb 2021 17:19:25 +0000 (12:19 -0500)]
Merge PR #39147 into master
* refs/pull/39147/head:
qa/tasks/ceph_fuse: do not createfs
qa/tasks/cephfs/fuse_mount: pass admin_socket path
qa/suites/fs/cephadm/multivolume: add basic multivolume test
mgr/mds_autoscaler: some fixes and cleanup
mgr/volumes: deploy MDSs when creating fs
Jason Dillaman [Thu, 4 Feb 2021 14:00:23 +0000 (09:00 -0500)]
qa/suites/rbd: drop require-osd-release command
Teuthology already defaults to quincy now and results in a failure
when trying to set to pacific. Additionally, drop the LUKS readbalance
test since it's unnecessary to duplicate that test.
Signed-off-by: Jason Dillaman <dillaman@redhat.com>
If the requested write length does not match the provided bufferlist
length, disable the move optimization and instead fallback to creating
a new sub-bufferlist for the object request.
Fixes: https://tracker.ceph.com/issues/49173 Signed-off-by: Jason Dillaman <dillaman@redhat.com>
Kefu Chai [Wed, 3 Feb 2021 17:34:44 +0000 (01:34 +0800)]
common/Timer: use mono_clock for clock_t
there is chance that the system clock is adjust by chrony or ntpd, so that
a timer scheduled at epoch time 12:34 would never get scheduled if
the system clock is changed to 12:35 when the system clock is still
12:33.
if we have event which schedules itself when it is fired, there is
chance that this event could be put in a black hole if the system clock
is adjusted in the way explained above. this could be a serious issue
if we rely on the timer to do critical things.
but we *might* suffer from this issue even if we use mono_clock for
clock_t in Timer, because in older libstdc++ and libc++,
condition_variable::wait_until() and condition_variable::wait_for() were
still using pthread_cond_timedwait(). this is fixed after glibc
introduced pthread_cond_clockwait() in v2.3.0. for more details, see
https://reviews.llvm.org/D65339 and
https://gcc.gnu.org/legacy-ml/gcc-patches/2019-09/msg00190.html
so we can schedule an event with real_clock::time_point. timer should
wait using a mono_time, but user might want to pass a
real_clock::time_point to it when, for instance, scheduling a task which
should be performed in future at a specified time. in this case, we need
to convert the real_clock::time_point to a "known" time_point.
Kefu Chai [Wed, 3 Feb 2021 16:33:36 +0000 (00:33 +0800)]
common/condition_variable_debug: support mono_clock
it's allowed by the C++ standard library, there is not reason that it's
forbidden by us.
and it's important to use mono_clock along with condition_variable to
address the issue when system lock is modified when we are waiting for
an event using condition_variable::wait_until() or
condition_variable::wait_for()
J. Eric Ivancich [Fri, 29 Jan 2021 17:03:50 +0000 (12:03 -0500)]
rgw: add rgw-gap-list-comparator tool
The rgw-gap-list tool can produce a number of false positives when the
cluster is being used during its run. One technique to minimize the
number of false positives is to run the tool twice and look for the
objects that appear in both lists. The rgw-gap-list-comparator tool is
designed to do this comparison.
J. Eric Ivancich [Thu, 17 Dec 2020 23:21:36 +0000 (18:21 -0500)]
rgw: add rgw-gap-list tool
Due to a prior bug (pr: 38228) tail rados objects of some RGW objects
could have been incorrectly deleted. This tool is designed to look for
such cases. It essentially does the opposite of rgw-orphan-list,
looking for rados objects that RGW expects to be there, but which are
not to be found.
IMPORTANT: This is very experimental at this point in time, and any
"results" produced should be verified by other means.
Signed-off-by: J. Eric Ivancich <ivancich@redhat.com> Signed-off-by: Michael Kidd <linuxkidd@gmail.com>
Casey Bodley [Wed, 3 Feb 2021 20:17:19 +0000 (15:17 -0500)]
cmake/rgw: forward spawn's compile options to rgw_common object library
since rgw_common is an OBJECT library, we can't use
target_link_libraries() for its dependency on spawn. we add its
include directories manually already with
$<TARGET_PROPERTY:spawn,INTERFACE_INCLUDE_DIRECTORIES>, but this didn't
pull in the compile definitions. this ultimately prevented the
WITH_BOOST_VALGRIND option from passing the BOOST_USE_VALGRIND
definition attached to boost::context
Sage Weil [Wed, 3 Feb 2021 15:38:49 +0000 (10:38 -0500)]
Merge PR #39069 into master
* refs/pull/39069/head:
mgr/cephadm/upgrade: tolerate pre-pacific upgrade state
mgr/cephadm/upgrade: scale down MDS cluster(s) for major version upgrades
mgr/cephadm: fix capitalization, level; drop elipses of log msgs
mgr/cephadm/upgrade: match against any repo_digest, not image_id
cephadm: return repo_digests (plural) in pull/inspect output
mgr/cephadm: include container_image_digests in inventory
cephadm: include image_digests list in 'ls' output
vstart.sh: only extract first container digest
mgr/cephadm: move release -> major translation to helper
mgr/cephadm/upgrade: tolerate old upgrade_state.target_versoin
mgr/cephadm/upgrade: set require-osd-release when done with OSDs
mgr: add lookup_release_name(int) to mgr interface
mgr/cephadm: verify container image version after we pull it
mgr/cephadm: only save version portion of version string
cephadm: fix 'inspect' and 'pull'
mgr/cephadm/upgrade: implement N-2 version checks on upgrade start
Casey Bodley [Wed, 3 Feb 2021 14:46:33 +0000 (09:46 -0500)]
cmake: partial revert of BOOST_USE_VALGRIND when ALLOCATOR=libc
the WITH_SYSTEM_BOOST binaries are not built with BOOST_USE_VALGRIND, so
it probably isn't safe to define for the headers only
this flag is needed for teuthology testing, and the shaman builds use
WITH_SYSTEM_BOOST=OFF. so the better fix is to enable WITH_BOOST_VALGRIND
so BuildBoost.cmake will build the libraries with valgrind support and add
-DBOOST_USE_VALGRIND to the necessary targets
this change was merged in https://github.com/ceph/ceph-build/pull/1736
Zac Dover [Wed, 3 Feb 2021 14:06:13 +0000 (00:06 +1000)]
doc/dev: Remove workbench mentions
This PR removes the "running-tests-in-cloud.rst"
file, which explains how to use ceph-workbench.
ceph-workbench is now deprecated, and the new
Teuthology documentation supplants the information
in the ceph-workbench-related documentation.
This PR also alters the "index.rst" file to remove
a link to "running-tests-in-cloud.rst".
Lucian Petrut [Wed, 3 Feb 2021 08:59:24 +0000 (08:59 +0000)]
win32*.sh: move debug symbols to separate files
This patch simplifies releasing Windows binaries along with debug
symbols.
By default, we're going to provide minimum debug information (-g1).
The symbols are extracted from the binaries and placed in separate
files in the ".debug" folder, which is used by gdb implicitly.
This is more convenient than having separate versions of the binaries,
with or without debug symbols.
projects / ceph.git / log