The ceph lockbox partition (part number 5) used with non lvm scenarios
and in non containerized deployment don't have a valid PARTUUID.
The value is set to 00000000-0000-0000-0000-000000000000 for each OSD
devices.
$ blkid -t PARTLABEL="ceph lockbox" -o value -s PARTUUID
00000000-0000-0000-0000-000000000000
00000000-0000-0000-0000-000000000000
00000000-0000-0000-0000-000000000000
00000000-0000-0000-0000-000000000000
00000000-0000-0000-0000-000000000000
When switching to containerized deployment we manually mount the lockbox
partition by using the PARTUUID.
Unfortunately because we have most of the time multiple OSD on the same
node we can't have the right symlink in /dev/disk/by-partuuid because it
will point to only one partition.
/dev/disk/by-partuuid/00000000-0000-0000-0000-000000000000 -> ../../sdb5
After the switch_to_containers playbook then only one OSD will restart
correctly and the other will try to access to the wrong device causing
error like 'xxxx is still in use'.
When deploying with containers and dmcrypt OSDs we force a PARTUUID
value during the ceph-disk prepare task.
Closes: https://bugzilla.redhat.com/show_bug.cgi?id=1616159
Signed-off-by: Dimitri Savineau <dsavinea@redhat.com>
When a container is already running on a non containerized node then the
umount ceph partition task is skipped.
This is due to the container ps command which always returns 0 even if
the filter matches nothing.
We should run the umount task when:
1/ the container command is failing (not installed) : rc != 0
2/ the container command reports running ceph-osd containers : rc == 0
Also we should not fail on the ceph directory listing.
Closes: https://bugzilla.redhat.com/show_bug.cgi?id=1616159
Signed-off-by: Dimitri Savineau <dsavinea@redhat.com>
(cherry picked from commit 39cfe0aa65)
When switching from a baremetal deployment to a containerized deployment
we only umount the OSD data partition.
If the OSD is encrypted (dmcrypt: true) then there's an additional
partition (part number 5) used for the lockbox and mount in the
/var/lib/ceph/osd-lockbox/ directory.
Because this partition isn't umount then the containerized OSD aren't
able to start. The partition is still mount by the system and can't be
remount from the container.
Closes: https://bugzilla.redhat.com/show_bug.cgi?id=1616159
Signed-off-by: Dimitri Savineau <dsavinea@redhat.com>
(cherry picked from commit 19edf707a5)
By changing the set ownership command from using the file module in combination with a with_items loop to a raw chown command, we can achieve a 98% performance increase here.
On a ceph cluster with a significant amount of directories and files in /var/lib/ceph, the file module has to run checks on ownership of all those directories and files to determine whether a change is needed.
In this case, we just want to explicitly set the ownership of all these directories and files to the ceph_uid
Added context note to all set proper ownership tasks
Signed-off-by: Kevin Jones <kevinjones@redhat.com>
(cherry picked from commit 47bf47c9d8)
Some people use the switch playbook to perform upgrade so they end up in
the same situation than https://bugzilla.redhat.com/show_bug.cgi?id=1650572
This is applying the same fix as
729744c6a8.
We don't want to fail on key that are not present since they will get
created after the mons are updated. They will be created by the task
"create potentially missing keys (rbd and rbd-mirror)".
Signed-off-by: Sébastien Han <seb@redhat.com>
sometimes we play the whole role `ceph-defaults` just to access the
default value of some variables. It means we play the `facts.yml` part
in this role while it's not desired. Splitting this role will speedup
the playbook.
Closes: #3282
Signed-off-by: Guillaume Abrioux <gabrioux@redhat.com>
(cherry picked from commit 0eb56e36f8)
It's easier lookup a directoriy instead of the block devices,
especially because of ceph-volume and ceph-disk have a different way to
handle devices.
Signed-off-by: Sébastien Han <seb@redhat.com>
(cherry picked from commit c14f9b78ff)
Prior to this commit we were only disabling ceph-osd units, but forgot
the ceph.target which is controlling everything and will restart the
ceph-osd units at each reboot.
Now that everything gets disabled there won't be any conflicts between
the old non-container and the new container units.
Signed-off-by: Sébastien Han <seb@redhat.com>
(cherry picked from commit cd56dad9fa)
If we mask it we won't be able to start the OSD container since now the
osd container use the osd ID as a name such as: ceph-osd@0
Fixes the error: Failed to execute operation: Cannot send after transport endpoint shutdown
Signed-off-by: Sébastien Han <seb@redhat.com>
(cherry picked from commit fe1d09925a)
The current regex had a limitation of 99 OSDs, now this limit has been
removed and regardless the number of OSDs they will all be collected.
Closes: https://bugzilla.redhat.com/show_bug.cgi?id=1630430
Signed-off-by: Sébastien Han <seb@redhat.com>
Fixes the deprecation warning:
[DEPRECATION WARNING]: Using tests as filters is deprecated. Instead of
using `result|search` use `result is search`.
Signed-off-by: Noah Watkins <nwatkins@redhat.com>
We need to copy this key into /etc/ceph so when ceph-docker-common runs
it can fetch it to the ansible server. Previously the task wasn't not
failing because `fail_on_missing` was False before 2.5, so now it's True
hence the failure.
Signed-off-by: Sébastien Han <seb@redhat.com>
Add missing call the ceph-handler role, otherwise we can't have
reference to variable registered from ceph-handler from other roles.
Signed-off-by: Guillaume Abrioux <gabrioux@redhat.com>
Similar to c13a3c3 we must allow scrubbing when running this playbook.
In cluster with a large number of PGs, it can be expected some of them
scrubbing, it's a normal operation.
Preventing from scrubbing operation force to set noscrub flag.
This commit allows to switch from non containerized to containerized
environment even while PGs are scrubbing.
Closes: #3182
Signed-off-by: Guillaume Abrioux <gabrioux@redhat.com>
During the transition from jewel non-container to container old ceph
units are disabled. ceph-disk can still remain in some cases and will
appear as 'loaded failed', this is not a problem although operators
might not like to see these units failing. That's why we remove them if
we find them.
Closes: https://bugzilla.redhat.com/show_bug.cgi?id=1577846
Signed-off-by: Sébastien Han <seb@redhat.com>
In addition to b324c17 this commit fix the ceph uid for osd role in the
switch from non containerized to containerized playbook.
Signed-off-by: Guillaume Abrioux <gabrioux@redhat.com>
If we don't do this, umounting devices declared like this
/dev/disk/by-id/ata-QEMU_HARDDISK_QM00001
will fail like:
umount: /dev/disk/by-id/ata-QEMU_HARDDISK_QM000011: mountpoint not found
Since we append '1' (partition 1), this won't work.
So we need to resolved the link to get something like /dev/sdb and then
append 1 to /dev/sdb1
Signed-off-by: Sébastien Han <seb@redhat.com>
Co-authored-by: Guillaume Abrioux <gabrioux@redhat.com>
We know bindmount with the :z option at the end of the -v command so
this will basically run the exact same command as we used to run. So to
speak:
chcon -Rt svirt_sandbox_file_t /var/lib/ceph
Signed-off-by: Sébastien Han <seb@redhat.com>
If we're working with a jewel cluster then this service will not exist.
This is mainly a problem with CI testing because our tests are setup to
work with both jewel and luminous, meaning that eventhough we want to
test jewel we still have a nfs-ganesha host in the test causing these
tasks to run.
Signed-off-by: Andrew Schoen <aschoen@redhat.com>
Use the pg check before doing the pg check, not on the quorum check.
Also never quote int when doing comparaison.
Signed-off-by: Sébastien Han <seb@redhat.com>
This commits allows us to run
switch-from-non-containerized-to-containerized-ceph-daemons.yml multiple
times.
Closes: https://bugzilla.redhat.com/show_bug.cgi?id=1489353
Signed-off-by: Sébastien Han <seb@redhat.com>
If devices is passed through an extra var this register won't work so
let's only register the var is devices is not defined.
Closes: https://bugzilla.redhat.com/show_bug.cgi?id=1489099
Signed-off-by: Sébastien Han <seb@redhat.com>
we need to force the value of `docker` variable which is initially set
to `false` since it's a migration from non-containerized to
containerized cluster.
Signed-off-by: Guillaume Abrioux <gabrioux@redhat.com>
We must mask the image so we are sure that even if the system reboots
then the OSDs won't start.
Also remove Ceph udev rules if found on the system prior to deploy
containers. If we don't do this we are exposed to conflicts between udev
rules and sytemd unit files.
Also add the CI will now test the migration from a non-containerized cluster to a
containerized cluster.
Signed-off-by: Sébastien Han <seb@redhat.com>
This commit eases the use of the
infrastructure-playbooks/switch-from-non-containerized-to-containerized-ceph-daemons.yml
playbook. We basically run it with a couple of pre-tasks and then we let
the playbook run the docker roles.
It obviously expect to have proper variables configured in order to
work.
Signed-off-by: Sébastien Han <seb@redhat.com>
In the switch to containers migration there were broken references
to ceph_mon_docker_subnet variable, replaced with public_network.
Also fixes references to ceph_mon_docker_extra_env setting for it
a default as it could be undefined.
remove `ceph_mon_docker_interface` and use `monitor_interface` instead
for both containerized and non-containerized deployment.
Signed-off-by: Guillaume Abrioux <gabrioux@redhat.com>
The new test in the checks PGs are no longer working on distributions
where /bin/sh isn't linked to /bin/bash.
Fix: #1619
Signed-off-by: Guillaume Abrioux <gabrioux@redhat.com>
Avoid screen scrapping by rewriting `waiting for clean pgs` tasks like it is
done in 304de48.
Use the json output returned by `ceph -s` instead
Signed-off-by: Guillaume Abrioux <gabrioux@redhat.com>
Problem: too many different commands to do the same thing. The 'cut'
command on infrastructure-playbooks/purge-cluster.yml was also wrong.
This sed command from osixia in ceph-docker
https://github.com/ceph/ceph-docker/pull/580/ addresses all the
scenarios.
Signed-off-by: Sébastien Han <seb@redhat.com>
Doing so will override any values set for these in the group_vars
directory relative to the users inventory.
Signed-off-by: Andrew Schoen <aschoen@redhat.com>
We changed the way we declare image.
Prior to this patch we must have a "user/image:tag"
format, which is incompatible with non docker-hub registry where you
usually don't have a "user". On the docker hub a "user" is also
identified as a namespace, so for Ceph the user was "ceph".
Variables have been simplified with only:
* ceph_docker_image
* ceph_docker_image_tag
1. For docker hub images: ceph_docker_name: "ceph/daemon" will give
you the 'daemon' image of the 'ceph' user.
2. For non docker hub images: ceph_docker_name: "daemon" will simply
give you the "daemon" image.
Infrastructure playbooks have been modified as well.
The file group_vars/all.docker.yml.sample has been removed as well.
It is hard to maintain since we have to generate it manually. If
you want to configure specific variables for a specific daemon simply
edit group_vars/$DAEMON.yml
Fixes: https://bugzilla.redhat.com/show_bug.cgi?id=1420207
Signed-off-by: Sébastien Han <seb@redhat.com>
This playbook was still referencing the old version of the
ceph_*_docker_extra_env but only for Ceph MONs and Ceph NFS. This
playbook was not kept up-to-date when updating the
ceph_*_docker_extra_env variables to add the '-e' option to docker.
That's because the addition of '-e' breaks this playbook as it requires
a comma separated list of variables for the 'env:' docker module
parameter. Therefore this change just makes the playbook consistently
broken by referencing the same variable throughout.
Dimitri Savineau
Guillaume Abrioux
Kevin Jones
Sébastien Han
Noah Watkins
Andrew Schoen
Giulio Fidente
Konstantin Shalygin
Tobias Florek
Ivan Font