centos/rhel have podman 2, which does not like conflicting --cap-add and
--privileged arguments. cephadm versions prior to 15.2.5 use both args,
however, which means the rhel/centos upgrade test has to start at 15.2.5
to work at all on those distros (with the updated podman).
Fixes: https://tracker.ceph.com/issues/48142
Signed-off-by: Sage Weil <sage@newdream.net>
When the following conditions are true:
1) A host has selinux-policy-targeted,
2) We mount the host's /sys into a privileged container,
3) The container has SELINUXTYPE=targeted in /etc/selinux/config,
4) The container does not have an selinux-policy-targeted package,
then SELinux-enabled applications like restorecon or DNF do not work inside
the container.
Resolve this by making /sys/fs/selinux an empty directory.
Fixes: https://tracker.ceph.com/issues/49239
Signed-off-by: Ken Dreyer <kdreyer@redhat.com>
doc/install: cephadm is the successor of DeepSea
Reviewed-by: Juan Miguel Olmo Martínez <jolmomar@redhat.com>
Reviewed-by: Michael Fritch <mfritch@suse.com>
crimson/osd: capture pgref by value if it is to be referenced asynchronously
Reviewed-by: Samuel Just <sjust@redhat.com>
Reviewed-by: Kefu Chai <kchai@redhat.com>
RocksDB's cmake has a similar logic to turn on ccache as the main cmake.
Currently BuildRocksDB.cmake explicitly passes ccache to RocksDB cmake
arguments as CMAKE_CXX_COMPILER_LAUNCHER. This causes a compilation error
on ccache 4.0, especially in make -j mode.
ccache: error: Recursive invocation (the name of the ccache binary must
be "ccache")
This commit lets cmake from RocksDB decide RULE_LAUNCH_COMPILE by itself.
Signed-off-by: Sheng Mao <shngmao@gmail.com>
This rewrites the "Getting Binaries" section
of the "Workflow" section of the "Tests:
Integration Tests" section of the Developer Guide.
Signed-off-by: Zac Dover <zac.dover@gmail.com>
the 'sudo netstat | grep radosgw' command is causing failures:
File "qa/workunits/rgw/test_rgw_reshard.py", line 53, in get_radosgw_port
x = out.decode('utf8').split(" ")
AttributeError: 'bool' object has no attribute 'decode'
because when valgrind is enabled, the process name is 'valgrind' instead
of 'radosgw'
Signed-off-by: Casey Bodley <cbodley@redhat.com>
If we make any change to the osd crush map like do an osd crush reweight from cli, for that change to be reflected on metadata table we need to reload the entire page. Instead this PR takes care of auto refreshing the tree view.
Fixes: https://tracker.ceph.com/issues/48922
Signed-off-by: Nizamudeen A <nia@redhat.com>
Signed-off-by: Avan Thakkar <athakkar@redhat.com>
Add a few notes about possible future scalability improvements
Co-authored-by: Joshua Schmid <jschmid@suse.de>
Signed-off-by: Sebastian Wagner <sebastian.wagner@suse.com>
This patch adds generic ETW logging on top of syslog calls.
The events are described in the message compiler file `event_logging.mc`.
Using the cross platform utility `windmc`, we will generate the header file,
`event_logging.h`, needed by the ETW implementation, and the resource file
`event_logging.rc`.
Over the generated resource file we will run another utility called `windres`.
this tool is used to compile the binary objects needed by the OS utilities
(i.e. `Event Viewer`) to view the logged events.
On usage, a registry key needs to be added/removed on the target computer.
The registry looks like the following:
[HKLM\SYSTEM\CurrentControlSet\Services\EventLog\Application\<get_process_name>]
"EventMessageFile"="<Folder_location_to>\\event_logging.dll"
I.E.:
[HKLM\SYSTEM\CurrentControlSet\Services\EventLog\Application\rbd-wnbd]
"EventMessageFile"="C:\\Program Files\\Ceph\\bin\\event_logging.dll"
Signed-off-by: Alin Gabriel Serdean <aserdean@cloudbasesolutions.com>
Signed-off-by: Lucian Petrut <lpetrut@cloudbasesolutions.com>
Co-authored-by: Lucian Petrut <lpetrut@cloudbasesolutions.com>
crimson/osd: do not pass oop as a const ref and cleanups
Reviewed-by: Radoslaw Zarzynski <rzarzyns@redhat.com>
Reviewed-by: Xuehan Xu <xxhdx1985126@gmail.com>
To be explicit that string_kew_view_t::MIN and MAX are not valid
string lengths. They are reserved as markers.
Signed-off-by: Yingxin Cheng <yingxin.cheng@intel.com>
structured binding does not define variables, let's capture the bound
"variables" using variables defined in the capture list.
Signed-off-by: Kefu Chai <kchai@redhat.com>
to silence -Wfinal-dtor-non-final-class warnings like
operation.h:461:20: warning: class with destructor marked 'final' cannot
be inherited from [-Wfinal-dtor-non-final-class]
~ExitBarrier() final {
^
Signed-off-by: Kefu Chai <kchai@redhat.com>
instead of using a local lambda which tries to have a single point
for handling both the happy path and failure path, it'd be simpler
to just early return if we fail to handle a request,
Signed-off-by: Kefu Chai <kchai@redhat.com>
before this change, std::move(oop) will create a copy of oop.
after this change, std::move(oop) actually moves oop.
Signed-off-by: Kefu Chai <kchai@redhat.com>
the order of evaluation of function arguments is unspecified with some
exceptions. but in this case, the reason why ops is still valid is that
the parameter type of `osd_op_p` is `const osd_op_params_t&`. so a copy
is passed to it.
to enable us to actually move `osd_op_p`, in this change, the
osd_op_p.req->ops is used instead.
Signed-off-by: Kefu Chai <kchai@redhat.com>
as we should expose the internals using the name of interface, and there
is no need to differentiate rep_repair_primary_object() from another
repair method.
Signed-off-by: Kefu Chai <kchai@redhat.com>
it errors while building without lua_packages:
+ -DBOOST_J=102 -DWITH_GRAFANA=ON
/var/tmp/rpm-tmp.aq2X3J: line 91: -DBOOST_J=102: command not found
Signed-off-by: luo.runbing <luo.runbing@zte.com.cn>
