keep the ceph.conf very simple.
manage the common options such as `public_network` with `ceph_config`
module.
Signed-off-by: Guillaume Abrioux <gabrioux@ibm.com>
This refactor makes the 'name' argument not mandatory because when
'state' is 'info' we shouldn't need to pass it.
The second change is just a duplicate code removal.
Signed-off-by: Guillaume Abrioux <gabrioux@ibm.com>
This adds the resquired changes in order to support
CentOS stream 9.
Also, this bumps the Ansible version support to 2.15
Signed-off-by: Guillaume Abrioux <gabrioux@ibm.com>
With ansible-core 2.15 it is not possible to pass argument of unexpected
type, as otherwise module will fail with:
`'None' is not a string and conversion is not allowed`
With that we want to only get all existing crush rules, so we can simply
supply an empty string as a name argument, which would satisfy
requirements and have same behaviour for previous ansible versions.
Alternative approach would be to stop making `name` as a required
argument to the module and use empty string as default value
when info state is used.
Signed-off-by: Dmitriy Rabotyagov <noonedeadpunk@gmail.com>
There was multiple rgw frontends entries while there was just one
rgw instance on each host. The other entries were the details from
the other rgw hosts in the cluster
Fixes: https://bugzilla.redhat.com/show_bug.cgi?id=2232282
Signed-off-by: Teoman ONAY <tonay@ibm.com>
When deploying with --skip-tags=package-install (when there is no access to a repository), the playbook is still trying to update the package cache, or to install ceph-mgr packages, which makes the playbook fail.
This change prevents the playbook to try to update the cache or install ceph-mgr packages when the package-install tag is skipped.
Signed-off-by: Florent CARLI <florent.carli@rte-france.com>
The tasks "manage nodes with cephadm - ipv4/6" are skipped when
cephadm_mgmt_network contains more than one ip network which prevent
cephadm from managing the host.
Signed-off-by: Teoman ONAY <tonay@ibm.com>
779523f86f introduced a regression
related to rbdmirrors tasks. They were executed while
ceph_rbd_mirror_remot_* variables were not set.
Signed-off-by: Teoman ONAY <tonay@ibm.com>
let's use quay.io/ceph/daemon-base in every tests instead of
`ceph/daemon` since it's not supposed to be built anymore soon.
Signed-off-by: Guillaume Abrioux <gabrioux@ibm.com>
As it's always being set in ceph.conf template, it leads to having duplicated osd_memory_target keys in rendered ceph conf while defining one in ceph_conf_overrides.
Signed-off-by: Seena Fallah <seenafallah@gmail.com>
This directory should be removed when the cluster is purged.
most of the services are started with the `--security-opt label=disable`
option. If the directory is not removed, it can cause SElinux issues
when the cluster is redeployed.
Signed-off-by: Guillaume Abrioux <gabrioux@redhat.com>
for some reason, this task has to be done in 2 steps otherwise it fails.
1/ stop and disable the service
2/ mask it
when done with with a single task, the module says the service has been
stopped while this isn't the case (Ansible systemd module bug?).
it possibly relates to https://github.com/ansible/ansible/issues/68680
Signed-off-by: Guillaume Abrioux <gabrioux@redhat.com>
filestore objectstore will be gone in the next Ceph release.the
This drops the filestore support in ceph-ansible.
Signed-off-by: Guillaume Abrioux <gabrioux@redhat.com>
`name` is a mandatory parameter when `state` is either
`present` or `absent` or `update` or `info`.
Signed-off-by: Guillaume Abrioux <gabrioux@redhat.com>
Instead of checking ip_version variable we should check the input
address for ip version and select code path based on that.
This solves ceph adoption with mixed ipv6 and ipv4 networks.
Resolves: rhbz#2186226
Signed-off-by: Lukas Bezdicka <lbezdick@redhat.com>
Exclude lvm_volumes defined disks from existing osds while it has been counted by the "count number of osds for lvm scenario" task.
Signed-off-by: Seena Fallah <seenafallah@gmail.com>
* Exclude device from lvm_volumes while osd_auto_discovery is true
* Sum num_osds on both lvm_volumes and devices list
Signed-off-by: Seena Fallah <seenafallah@gmail.com>
Enabling installation of the admin key to mgr nodes by setting
"copy_admin_key: true" is broken. This is because the variable is not
referenced correctly (using inline Jinja2 templating).
Signed-off-by: René Højbjerg Larsen <rhl@jfm.dk>
Add --security-opt label=disable to all containers
accessing /var/lib/ceph. podman selinux relabeling behavious changed
since version podman-3:4.2.0-1 which prevent some containers to access
files in these subdirectories.
Signed-off-by: Teoman ONAY <tonay@ibm.com>
If a disk has a symlink it will be re-added to the devices lists one with resolved path and the other with a defined path.
We can rebuild the list from the readlink output cause readlink always return the correct path for all disks.
Signed-off-by: Seena Fallah <seenafallah@gmail.com>
Exclude disks were defined in dedicated_devices and bluestore_wal_devices on osd_auto_discovery enabled.
Signed-off-by: Seena Fallah <seenafallah@gmail.com>
With podman version podman-3:4.2.0-4.module+el8.7.0+17064+3b31f55c and
later, when mgr fails to start if mon is already running.
Closes: https://bugzilla.redhat.com/show_bug.cgi?id=2169767
Signed-off-by: Teoman ONAY <tonay@ibm.com>
We need to make sure `rgw_instances` is set before `ceph.conf` is
rendered. Otherwise, the `ceph-crash` play in the main playbook updates
(via ceph-handler) the `ceph.conf` on rgw nodes and removes rgw instances
sections.
Closes: https://bugzilla.redhat.com/show_bug.cgi?id=2141604
Signed-off-by: Guillaume Abrioux <gabrioux@redhat.com>
Guillaume Abrioux
Guillaume Abrioux
Dmitriy Rabotyagov
Teoman ONAY
insatomcat
Nathan Hoad
Konstantin Shalygin
Seena Fallah
Lukas Bezdicka
René Højbjerg Larsen
Lorenz Bausch