The dashboard nodes (alertmanager, grafana, node-exporter, and prometheus)
were not manage during the docker to podman migration.
This adds the systemd container template of those services to a dedicated
file (systemd.yml) in order to include it in the docker2podman playbook.
This also adds the dashboard container images pull from docker to podman.
Closes: https://bugzilla.redhat.com/show_bug.cgi?id=1829389
Signed-off-by: Dimitri Savineau <dsavinea@redhat.com>
The docker2podman playbook only installs the podman package and updates
the systemd units with the right container_binary value.
We never pull the container image so if one service is restarted then
the container image will be pulled first before the service can start
which could cause longer downstream.
To avoid to download the container image from internet again we can just
pull it from the local docker daemon.
The container_{binding,package,service}_name variables are removed
because they are only used in the ceph-container-engine role which
isn't call in this playbook.
Signed-off-by: Dimitri Savineau <dsavinea@redhat.com>
When using skipped variables with from_json filter and python2 then we
need to have a default value otherwise the skipped task will fail.
Unexpected templating type error occurred on
({{ (ceph_volume_lvm_list.stdout | from_json) }}): expected string or
buffer
Closes: https://bugzilla.redhat.com/show_bug.cgi?id=1790472
Signed-off-by: Dimitri Savineau <dsavinea@redhat.com>
Since [1] we need to set pacific for the required OSD release during the
upgrade.
[1] https://github.com/ceph/ceph/commit/cc99c3bc
Signed-off-by: Dimitri Savineau <dsavinea@redhat.com>
The workflow in this playbook should be the same than in rolling_update,
we should first set noout and nodeep-scrub flags before migrating the
first osd and unset osd flags after the last osd is migrated.
Signed-off-by: Guillaume Abrioux <gabrioux@redhat.com>
When setting/unsetting osd flags, we can use `tasks_from` when importing
`ceph-facts` role to save some times given that we only need this role
for setting `container_binary`
Signed-off-by: Guillaume Abrioux <gabrioux@redhat.com>
We must call `ceph-osd` role from `container_options_facts.yml` because
ceph-osd-run.sh.j2 needs variables set in this file.
Closes: https://bugzilla.redhat.com/show_bug.cgi?id=1819681
Signed-off-by: Guillaume Abrioux <gabrioux@redhat.com>
This commits removes these two symlinks.
They were there for backward compatibility and were marked deprecated as
of stable-4.0
Signed-off-by: Guillaume Abrioux <gabrioux@redhat.com>
Adding `--all` to the `systemctl list-units` command in order to get
*all* osds id on the node (including stoppped osds). Otherwise, it will
purge the cluster but there will be leftover after that.
Closes: https://bugzilla.redhat.com/show_bug.cgi?id=1814542
Signed-off-by: Guillaume Abrioux <gabrioux@redhat.com>
Since Ceph Octopus is python3 only we don't need to specify the max open
files anymore with the container engine.
Signed-off-by: Dimitri Savineau <dsavinea@redhat.com>
We only disable the ceph-osd services but not the ceph-volume lvm
services during the filestore to bluestore migration.
Signed-off-by: Dimitri Savineau <dsavinea@redhat.com>
We should add retry/delay to check the presence of the rbdmirror daemon
in the cluster status because the status takes some time to be updated.
Also the metadata.hostname isn't a good key to check because it doesn't
reflect the ansible_hostname fact. We should use metadata.id instead.
Signed-off-by: Dimitri Savineau <dsavinea@redhat.com>
This playbook was using mds systemd condition.
Also a command task was using pipeline which is not allowed.
Signed-off-by: Dimitri Savineau <dsavinea@redhat.com>
The ceph-facts are running on localhost so if this node is using a
different OS/release that the ceph node we can have a mismatch between
docker/podman container binary.
This commit also reduces the scope of the ceph-facts role because we only
need the container_binary tasks.
Signed-off-by: Dimitri Savineau <dsavinea@redhat.com>
It looks like that the service module doesn't support wildcard anymore
for stopping/disabling multiple services.
fatal: [rgw0]: FAILED! => changed=false
msg: 'This module does not currently support using glob patterns,
found ''*'' in service name: ceph-radosgw@*'
...ignoring
Instead we should iterate over the rgw_instances list.
Signed-off-by: Dimitri Savineau <dsavinea@redhat.com>
When running environment with OSDs having ID with more than 2 digits,
some tasks don't match the system units and therefore, playbook can fail.
Closes: https://bugzilla.redhat.com/show_bug.cgi?id=1805643
Signed-off-by: Guillaume Abrioux <gabrioux@redhat.com>
We need to call ceph-facts only for setting `container_binary`.
Since this task has been isolated we can use `tasks_from` to only execute the
needed task.
Signed-off-by: Guillaume Abrioux <gabrioux@redhat.com>
If the filestore configuration was using a dedicated journal with either
a partition or a LV/VG then we need to reuse this for bluestore DB.
When filestore is using a raw devices then we shouldn't destroy
everything (data + journal) but only data otherwise the journal
partition won't exist anymore.
Closes: https://bugzilla.redhat.com/show_bug.cgi?id=1790479
Signed-off-by: Dimitri Savineau <dsavinea@redhat.com>
Instead of running shring-osd tasks on localhost and delegating most of
them to the first monitor, run all of them on the first monitor
directly.
This has the added advantage of becoming root on the monitor only, not
on localhost.
Signed-off-by: Benoît Knecht <bknecht@protonmail.ch>
We only support python3 so renaming all ceph python packages.
Some ceph packages were missing from the list (ceph-mon, ceph-osd or
rbd-mirror) or didn't exist anymore (ceph-fs-common, libcephfs1).
Signed-off-by: Dimitri Savineau <dsavinea@redhat.com>
This commit increases the default values for the following variable
consumed in switch-from-non-containerized-to-containerized-ceph-daemons.yml
playbook.
This also moves these variables in `ceph-defaults` role so the user can
set different values if needed.
Closes: https://bugzilla.redhat.com/show_bug.cgi?id=1783223
Signed-off-by: Guillaume Abrioux <gabrioux@redhat.com>
Fix purge cluster failed when local container images does not exist.
Purge node-exporter and grafana-server only when dashboard_enabled is set to True.
Signed-off-by: wujie1993 qq594jj@gmail.com
If the playbook is used on a host running bluestore OSDs then the
osd_fsid_list won't be filled because the bluestore OSDs are reported
with 'type: block' via ceph-volume lvm list command but we are looking
for 'type: data' (filestore).
TASK [zap ceph-volume prepared OSDs] *********
fatal: [xxxxx]: FAILED! =>
msg: '''osd_fsid_list'' is undefined
Closes: https://bugzilla.redhat.com/show_bug.cgi?id=1729267
Signed-off-by: Dimitri Savineau <dsavinea@redhat.com>
If the OSD node is already using bluestore OSDs then we should skip
all the remaining tasks to avoid purging OSD for nothing.
Instead we warn the user.
Closes: https://bugzilla.redhat.com/show_bug.cgi?id=1790472
Signed-off-by: Dimitri Savineau <dsavinea@redhat.com>
When the PV is already removed from the devices then we should not fail
to avoid errors like:
stderr: No PV found on device /dev/sdb.
Closes: https://bugzilla.redhat.com/show_bug.cgi?id=1729267
Signed-off-by: Dimitri Savineau <dsavinea@redhat.com>
When upgrading from RHCS 3.x where ceph-metrics was deployed on a
dedicated node to RHCS 4.0, it fails like following:
```
fatal: [magna005]: FAILED! => changed=false
gid: 0
group: root
mode: '0755'
msg: 'chown failed: failed to look up user ceph'
owner: root
path: /etc/ceph
secontext: unconfined_u:object_r:etc_t:s0
size: 4096
state: directory
uid: 0
```
because we are trying to run `ceph-config` on this node, it doesn't make
sense so we should simply run this play on all groups except
`[grafana-server]`.
Closes: https://bugzilla.redhat.com/show_bug.cgi?id=1793885
Signed-off-by: Guillaume Abrioux <gabrioux@redhat.com>
When osd_auto_discovery is set then we need to refresh the
ansible_devices fact between after the filestore OSD purge
otherwise the devices fact won't be populated.
Also remove the gpt header on ceph_disk_osds_devices because
the devices is empty at this point for osd_auto_discovery.
Adding the bool filter when needed.
Closes: https://bugzilla.redhat.com/show_bug.cgi?id=1729267
Signed-off-by: Dimitri Savineau <dsavinea@redhat.com>
We still need --destroy when using a raw device otherwise we won't be
able to recreate the lvm stack on that device with bluestore.
Running command: /usr/sbin/vgcreate -s 1G --force --yes ceph-bdc67a84-894a-4687-b43f-bcd76317580a /dev/sdd
stderr: Physical volume '/dev/sdd' is already in volume group 'ceph-b7801d50-e827-4857-95ec-3291ad6f0151'
Unable to add physical volume '/dev/sdd' to volume group 'ceph-b7801d50-e827-4857-95ec-3291ad6f0151'
/dev/sdd: physical volume not initialized.
--> Was unable to complete a new OSD, will rollback changes
Closes: https://bugzilla.redhat.com/show_bug.cgi?id=1792227
Signed-off-by: Dimitri Savineau <dsavinea@redhat.com>
When using docker 1.13.1, the current condition:
```
{% if (container_binary == 'docker' and ceph_docker_version.split('.')[0] is version_compare('13', '>=')) or container_binary == 'podman' -%}
```
is wrong because it compares the first digit (1) whereas it should
compare the second one.
It means we always use `--cpu-quota` although documentation recommend
using `--cpus` when docker version is 1.13.1 or higher.
From the doc:
> --cpu-quota=<value> Impose a CPU CFS quota on the container. The number of
> microseconds per --cpu-period that the container is limited to before
> throttled. As such acting as the effective ceiling.
> If you use Docker 1.13 or higher, use --cpus instead.
Signed-off-by: Guillaume Abrioux <gabrioux@redhat.com>
the new ceph status registered in `ceph_status` will report `fsmap.up` =
0 when it's the last mds given that it's done after we shrink the mds,
it means the condition is wrong. Also adding a condition so we don't try
to delete the fs if a standby node is going to rejoin the cluster.
Closes: https://bugzilla.redhat.com/show_bug.cgi?id=1787543
Signed-off-by: Guillaume Abrioux <gabrioux@redhat.com>
This commit lets add-osd.yml in place but mark the deprecation of the
playbook.
Scaling up OSDs is now possible using --limit
Signed-off-by: Guillaume Abrioux <gabrioux@redhat.com>
play vars have lower precedence than role vars and `set_fact`.
We must use a `set_fact` to reset these variables.
Signed-off-by: Guillaume Abrioux <gabrioux@redhat.com>
This commit adds a package installation task in order to install podman
during the docker-to-podman.yml migration playbook.
Signed-off-by: Guillaume Abrioux <gabrioux@redhat.com>
We only need to have the container_binary fact. Because we're not
gathering the facts from all nodes then the purge fails trying to get
one of the grafana fact.
Closes: https://bugzilla.redhat.com/show_bug.cgi?id=1786686
Signed-off-by: Dimitri Savineau <dsavinea@redhat.com>
There's some tasks using the new container image during the rolling
upgrade playbook that needs to execute the registry login first otherwise
the nodes won't be able to pull the container image.
Unable to find image 'xxx.io/foo/bar:latest' locally
Trying to pull repository xxx.io/foo/bar ...
/usr/bin/docker-current: Get https://xxx.io/v2/foo/bar/manifests/latest:
unauthorized
Signed-off-by: Dimitri Savineau <dsavinea@redhat.com>
Instead of running the ceph roles against localhost we should do it
on the first mon.
The ansible and inventory hostname of the rgw nodes could be different.
Ensure that the rgw instance to remove is present in the cluster.
Fix rgw service and directory path.
Closes: https://bugzilla.redhat.com/show_bug.cgi?id=1677431
Signed-off-by: Dimitri Savineau <dsavinea@redhat.com>
The command is delegated on the first monitor so we must use the fact
`container_binary` from this node.
Signed-off-by: Guillaume Abrioux <gabrioux@redhat.com>
This commit deletes the filesystem when no more MDS is present after
shrinking operation.
Closes: https://bugzilla.redhat.com/show_bug.cgi?id=1787543
Signed-off-by: Guillaume Abrioux <gabrioux@redhat.com>
This commit prevent from shrinking an mds node when max_mds wouldn't be
honored after that operation.
Signed-off-by: Guillaume Abrioux <gabrioux@redhat.com>
When using the ceph dashboard with iscsi gateways nodes we also need to
remove the nodes from the ceph dashboard list.
Closes: https://bugzilla.redhat.com/show_bug.cgi?id=1786686
Signed-off-by: Dimitri Savineau <dsavinea@redhat.com>
on master we can't test upgrade from stable-4.0/CentOS 7 to
master/CentOS 8.
This commit refact the upgrade so we test upgrade from master/CentOS 8
to master/CentOS 8 (octopus to octopus)
Signed-off-by: Guillaume Abrioux <gabrioux@redhat.com>
When an OSD is stopped, it leaves partitions mounted.
We must umount them before zapping them, otherwise error like "Device is
busy" will show up.
Closes: https://bugzilla.redhat.com/show_bug.cgi?id=1729267
Signed-off-by: Guillaume Abrioux <gabrioux@redhat.com>
This commit adds a task to ensure device mappers are well closed when
lvm batch scenario is used.
Otherwise, OSDs can't be redeployed given that devices that are rejected
by ceph-volume because they are locked.
Adding a condition `devices | default([]) | length > 0` to remove these
dm only when using lvm batch scenario.
Signed-off-by: Guillaume Abrioux <gabrioux@redhat.com>
Otherwise, sometimes it can take a while for an OSD to be seen as down
and causes the `ceph osd purge` command to fail.
Signed-off-by: Guillaume Abrioux <gabrioux@redhat.com>
Do not use `--destroy` when zapping a device.
Otherwise, it destroys VGs while they are still needed to redeploy the
OSDs.
Signed-off-by: Guillaume Abrioux <gabrioux@redhat.com>
This commit adds the non containerized context support to the
filestore-to-bluestore.yml infrastructure playbook.
Closes: https://bugzilla.redhat.com/show_bug.cgi?id=1729267
Signed-off-by: Guillaume Abrioux <gabrioux@redhat.com>
When using fqdn in inventory, that playbook fails because of some tasks
using the result of ceph osd tree (which returns shortname) to get
some datas in hostvars[].
Closes: https://bugzilla.redhat.com/show_bug.cgi?id=1779021
Signed-off-by: Guillaume Abrioux <gabrioux@redhat.com>
In containerized context, containers aren't stopped early in the
sequence.
It means they aren't restarted after the upgrade because the task is
just checking the daemon status is started (eg: `state: started`).
This commit also removes the task which ensure services are started
because it's already done in the role ceph-iscsigw.
Signed-off-by: Guillaume Abrioux <gabrioux@redhat.com>
when upgrading from RHCS 3, dashboard has obviously never been deployed
and it forces us to deploy it later manually.
This commit adds the dashboard deployment as part of the upgrade to
RHCS 4.
Closes: https://bugzilla.redhat.com/show_bug.cgi?id=1779092
Signed-off-by: Guillaume Abrioux <gabrioux@redhat.com>
The podman support was added to the purge-container-cluster playbook but
containers are always used for the dashboard even on non containerized
deployment.
This commits adds the podman support on purging the dashboard resources
in the purge-cluster playbook.
Signed-off-by: Dimitri Savineau <dsavinea@redhat.com>
ceph/ceph-ansible#4805 introduced a symlink to
purge-container-cluster.yml playbook which is broken.
This commit fixes it.
Signed-off-by: Guillaume Abrioux <gabrioux@redhat.com>
All containers are removed when systemd stops them.
There is no need to call this module in purge container playbook.
This commit also removes all docker_image task and remove all container
images in the final cleanup play.
Closes: https://bugzilla.redhat.com/show_bug.cgi?id=1776736
Signed-off-by: Guillaume Abrioux <gabrioux@redhat.com>
This is needed to avoid following error:
```
ERROR! The requested handler 'restart ceph mons' was not found in either the main handlers list nor in the listening handlers list
```
Closes: https://bugzilla.redhat.com/show_bug.cgi?id=1777829
Signed-off-by: Guillaume Abrioux <gabrioux@redhat.com>
We must import this role in the first play otherwise the first call to
`client_group_name`fails.
Closes: https://bugzilla.redhat.com/show_bug.cgi?id=1777829
Signed-off-by: Guillaume Abrioux <gabrioux@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>
1. set noout and nodeep-scrub flags,
2. upgrade each OSD node, one by one, wait for active+clean pgs
3. after all osd nodes are upgraded, unset flags
Signed-off-by: Guillaume Abrioux <gabrioux@redhat.com>
Co-authored-by: Rachana Patel <racpatel@redhat.com>
It might be possible at some point even with osd flags `noout` and
`norebalance` set the PGs states can change depending on the amount of data
written meantime. It means the check for PGs state will fail.
This commit changes the way we set those flags:
we set them before an OSD node upgrade and unset them before the PGs
state check so they can recover.
Fixes: #3961
Signed-off-by: Guillaume Abrioux <gabrioux@redhat.com>
in containerized context, using the binary provided in atomic os won't
work because it's an old version provided by ceph-common based on
10.2.5.
Using a container could be an idea but for large cluster with hundreds
of client nodes, that would require to pull the image of each of them
just to unmap the rbd devices.
Let's use the sysfs method in order to avoid any issue related to ceph
version that is shipped on the host.
Closes: https://bugzilla.redhat.com/show_bug.cgi?id=1766064
Signed-off-by: Guillaume Abrioux <gabrioux@redhat.com>
If the new mon/osd node doesn't have python installed then we need to
execute the tasks from raw_install_python.yml.
Closes: #4368
Signed-off-by: Dimitri Savineau <dsavinea@redhat.com>
This commit adds a default value in the `with_dict` because when using
python 2.7, if a task using a `with_dict` has a condition, it is
evaluated anyway whereas in python 3 it isn't.
Closes: https://bugzilla.redhat.com/show_bug.cgi?id=1766499
Signed-off-by: Guillaume Abrioux <gabrioux@redhat.com>
There's no need to use the default filter on active/standby groups
because if the group doesn't exist then the play is just skipped.
Currently this generates warnings like:
[WARNING]: Could not match supplied host pattern, ignoring: |
[WARNING]: Could not match supplied host pattern, ignoring: default([])
Signed-off-by: Dimitri Savineau <dsavinea@redhat.com>
The active mds host should be based on the inventory hostname and not on
the ansible hostname.
The value returns under the mdsmap structure is based on the OS hostname
so we need to find the right node in the inventory with this value when
doing operation on inventory nodes.
Othewise we could see error like:
The task includes an option with an undefined variable. The error was:
"hostvars[foobar]" is undefined
Signed-off-by: Dimitri Savineau <dsavinea@redhat.com>
mon_host should use the inventory hostname and not the node hostname.
Fix creates an issue when the inventory and node hostname are different.
Closes: #4670
Signed-off-by: Dimitri Savineau <dsavinea@redhat.com>
The ceph-container-engine role is missing from both playbooks so the
container engine (docker, podman) isn't install resulting in a failure
on the added nodes.
fatal: [xxxxx]: FAILED! => changed=false
cmd: docker --version
msg: '[Errno 2] No such file or directory'
rc: 2
Closes: #4634
Signed-off-by: Dimitri Savineau <dsavinea@redhat.com>
The [group|host]_vars directories are ignored for the dashboard playbook
when the inventory file directory doesn't contain those directories.
Closes: #4601
Closes: https://bugzilla.redhat.com/show_bug.cgi?id=1761612
Signed-off-by: Dimitri Savineau <dsavinea@redhat.com>
The common roles don't need to be executed again on each group plays
(like mons, osds, etc..).
We only need to execute them during the first play. That wat, we will
apply the changes on all nodes in parallel instead of doing it once per
group.
Signed-off-by: Dimitri Savineau <dsavinea@redhat.com>
This is already done in the main playbooks but absent in the dashboard
playbook.
The facts are already gathered during the first play of the main
playbooks so we don't need to doing twice.
Signed-off-by: Dimitri Savineau <dsavinea@redhat.com>
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>
This commit refacts the way we set `ceph_uid` fact in `ceph-facts` and
removes all `set_fact` tasks for `ceph_uid` in switch-to-containers playbook
to avoid duplicated code.
Signed-off-by: Guillaume Abrioux <gabrioux@redhat.com>
This commit excludes client nodes from facts gathering, they are not
needed and can speed up this task.
Signed-off-by: Guillaume Abrioux <gabrioux@redhat.com>
The block section were used with the dashboard_enabled condition when
the code was included in the main playbooks.
Because this condition isn't present in the dashboard playbook anymore
we can remove the block section.
Signed-off-by: Dimitri Savineau <dsavinea@redhat.com>
This commit moves containerized deployment related files to `./tasks/`
directory. This is needed to make `docker-to-podman.yml` working since
we use `tasks_from:` option.
Signed-off-by: Guillaume Abrioux <gabrioux@redhat.com>
This commit moves containerized deployment related files to `./tasks/`
directory. This is needed to make `docker-to-podman.yml` working since
we use `tasks_from:` option.
Signed-off-by: Guillaume Abrioux <gabrioux@redhat.com>
Guillaume Abrioux
Dimitri Savineau
Benoît Knecht
wujie1993