This commit changes the bind mount option for the mount point
`/var/lib/ceph` in the systemd template for mon and mgr containers. This
is needed in case of collocating mon/mgr with osds using dmcrypt
scenario.
Once mon/mgr got converted to containers, the dmcrypt layer sub mount is
still seen in `/var/lib/ceph`. For some reason it makes the
corresponding devices busy so any other container can't open/close it.
As a result, it prevents osds from starting properly.
Since it only happens on the nodes converted before the OSD play, the idea is
to bind mount `/var/lib/ceph` on mon and mgr with the `rshared` option
so once the sub mount is unmounted, it is propagated inside the
container so it doesn't see that mount point.
Closes: https://bugzilla.redhat.com/show_bug.cgi?id=1896392
Signed-off-by: Guillaume Abrioux <gabrioux@redhat.com>
(cherry picked from commit f5ba6d9b01)
when using `monitor_interface`, if nodes don't have same interface names
this task will fail like following:
```
fatal: [argo010]: FAILED! => {
"msg": "The task includes an option with an undefined variable. The error was: 'ansible.vars.hostvars.HostVarsVars object' has no attribute u'ansible_enp1s0f0'\n\nThe error appears to have been in '/usr/share/ceph-ansible/roles/ceph-mon/tasks/docker/main.yml': line 19, column 3, but may\nbe elsewhere in the file depending on the exact syntax problem.\n\nThe offending line appears to be:\n\n\n- name: ipv4 - force peer addition as potential bootstrap peer for cluster bringup - monitor_interface\n ^ here\n"
}
```
Closes: https://bugzilla.redhat.com/show_bug.cgi?id=1876551
Signed-off-by: Guillaume Abrioux <gabrioux@redhat.com>
90f3f61 introduced the docker-to-podman.yml playbook but the
ceph-osd-run.sh.j2 template still has some docker hardcoded instead
of using the container_binary variable.
Signed-off-by: Dimitri Savineau <dsavinea@redhat.com>
This together with condition 'ansible_play_hosts_all | last' causes skipping that task on the first host.
Signed-off-by: RPietrzak <rp.pietrzak@gmail.com>
This commit moves the systemd rendering task into `systemd.yml` file.
Otherwise, when running docker to podman playbook, the systemd unit file
isn't updated as it should be.
Closes: https://bugzilla.redhat.com/show_bug.cgi?id=1870141
Signed-off-by: Guillaume Abrioux <gabrioux@redhat.com>
This commit makes the bindmount a bit more generic, otherwise it
currently makes the OSDs failing to start in an OSP FFU upgrade
(with RHEL7 > RHEL8 OS upgrade).
docker2podman playbook is run from ceph-ansible stable-3.2 branch
against RHEL7 nodes where `/var/run/lvmetad.socket` exists but once the
system is upgraded to RHEL8, this socket doesn't exist anymore and
prevent OSDs from starting after the reboot.
As a workaround we can make this bindmount a bit more generic like what
is done in `stable-4.0` branch by mounting `/run/lvm` instead.
Closes: https://bugzilla.redhat.com/show_bug.cgi?id=1866252
Signed-off-by: Guillaume Abrioux <gabrioux@redhat.com>
When using collocated or non-collocated osd_scenarios (ceph-disk) and
trying to deterime the OSD_DEVICE from the OSD_ID passed to the systemd
unit then we can be in a situation where the OSD hasn't been activated
but the OSD ID exists.
This means the data partition isn't in activate state and the ceph-disk
list command won't show the OSD ID on the data partition.
Closes: https://bugzilla.redhat.com/show_bug.cgi?id=1850377
Signed-off-by: Dimitri Savineau <dsavinea@redhat.com>
This isn't backported from master because there are too many changes
between stable-3.2 and other newer branches.
NOTE:
This playbook *doesn't* add podman support in stable-3.2 at all.
This is a tripleO dedicated playbook which is intended to be run
early during FFU workflow in order to prepare the OS upgrade.
Closes: https://bugzilla.redhat.com/show_bug.cgi?id=1853457
Signed-off-by: Guillaume Abrioux <gabrioux@redhat.com>
We shouldn't set this flag when running switch_to_containers playbook.
Otherwise the playbook fails waiting for pgs to be clean.
Closes: https://bugzilla.redhat.com/show_bug.cgi?id=1843569
Signed-off-by: Guillaume Abrioux <gabrioux@redhat.com>
(cherry picked from commit b91d60d384)
When using docker container engine then the systemd unit scripts only
use a dependency on the docker daemon via the After parameter.
But if docker is restarted on a live system then the ceph systemd units
should wait for the docker daemon to be fully restarted.
Closes: https://bugzilla.redhat.com/show_bug.cgi?id=1846830
Signed-off-by: Dimitri Savineau <dsavinea@redhat.com>
(cherry picked from commit bd22f1d1ec)
We shouldn't set this flag when running switch_to_containers playbook.
Otherwise the playbook fails waiting for pgs to be clean.
Closes: https://bugzilla.redhat.com/show_bug.cgi?id=1843569
Signed-off-by: Guillaume Abrioux <gabrioux@redhat.com>
(cherry picked from commit b91d60d384)
This commit moves the dummy container creation task right before the
cephx keys creation task so it can't be run out of time.
Also, this commit makes the dummy container running for ever.
Closes: https://bugzilla.redhat.com/show_bug.cgi?id=1828105
Signed-off-by: Guillaume Abrioux <gabrioux@redhat.com>
We were not testing the right ansible_distribution fact value for RHEL
distribution.
This commit also updates the minial RHEL version supported by RHCS.
Signed-off-by: Dimitri Savineau <dsavinea@redhat.com>
(cherry picked from commit 5de74fe512)
this commit fixes a bug when using `add-osd.yml` playbook.
`noup` flag is set early but it never got unset before the "wait for pgs
clean" check, so the playbook always fails because OSDs aren't never
seen UP.
Closes: https://bugzilla.redhat.com/show_bug.cgi?id=1816023
Signed-off-by: Guillaume Abrioux <gabrioux@redhat.com>
Creating crush rules even with no crush hierarchy configuration is a
valid scenario so we shouldn't be bound to the first task result (which
configure crush hierarchy) to be able to add new crush rules.
Closes: https://bugzilla.redhat.com/show_bug.cgi?id=1816989
Signed-off-by: Guillaume Abrioux <gabrioux@redhat.com>
(cherry picked from commit 5b0476385c)
Client configuration with --limit fails without this patch
because certain tasks are only done to the first host in the
_filtered_clients list and it's likely that first host will
not be included in what's sepcified with --limit. To fix this
the _filtered_clients list should be built from all clients
in the inventory that are also in the running play.
Closes: https://bugzilla.redhat.com/show_bug.cgi?id=1798781
Signed-off-by: John Fulton <fulton@redhat.com>
(cherry picked from commit e4bf4857f5)
This is no longer true, let's remove this comment given that this option
is not ignored in containerized deployments.
Signed-off-by: Guillaume Abrioux <gabrioux@redhat.com>
(cherry picked from commit e551b5ba1a)
Those tasks aren't needed in docker-common since the introduction of
`ceph-infra` role. They are duplicated tasks.
Closes: https://bugzilla.redhat.com/show_bug.cgi?id=1810376
Signed-off-by: Guillaume Abrioux <gabrioux@redhat.com>
(cherry picked from commit cd0195c562)
In 3c31b19ab3, I fixed the `customize pool
size` task by replacing `item.size` with `item.value.size`. However, I
missed the same issue in the `when` condition.
Signed-off-by: Benoît Knecht <bknecht@protonmail.ch>
(cherry picked from commit 3842aa1a30)
RadosGW pools can be created by setting
```yaml
rgw_create_pools:
.rgw.root:
pg_num: 512
size: 2
```
for instance. However, doing so would create pools of size
`osd_pool_default_size` regardless of the `size` value. This was due to
the fact that the Ansible task used
```
{{ item.size | default(osd_pool_default_size) }}
```
as the pool size value, but `item.size` is always undefined; the
correct variable is `item.value.size`.
Signed-off-by: Benoît Knecht <bknecht@protonmail.ch>
(cherry picked from commit 3c31b19ab3)
Since ed36a11 we move the crush rules creation code from the ceph-mon to
the ceph-osd role.
To keep the backward compatibility we kept the possibility to set the
crush variables on the mons side but we didn't move the default values.
As a result, when using crush_rule_config set to true and wanted to use
the default values for crush_rules then the crush rule ansible task
creation will fail.
"msg": "'ansible.vars.hostvars.HostVarsVars object' has no attribute
'crush_rules'"
This patch move the default crush variables from ceph-mon to ceph-osd
role but also use those default values when nothing is defined on the
mons side.
Closes: https://bugzilla.redhat.com/show_bug.cgi?id=1798864
Signed-off-by: Dimitri Savineau <dsavinea@redhat.com>
(cherry picked from commit 1fc6b33714)
We already stop the upgrade playbook (rolling_update.yml) if there's
no mgr node present so we should also do the same for initial
deployment.
Closes: https://bugzilla.redhat.com/show_bug.cgi?id=1788644
Signed-off-by: Dimitri Savineau <dsavinea@redhat.com>
When using ceph aliases with commands that require manual intervention
to stop then the command will keep running inside the container (like
using Ctrl+c).
For handling this, we should use the interactive session option (-it)
with the docker commands.
Closes: https://bugzilla.redhat.com/show_bug.cgi?id=1797874
Signed-off-by: Dimitri Savineau <dsavinea@redhat.com>
During a rolling update we will run the ceph iscsigw tasks that start
the daemons then run the configure_iscsi.yml tasks which can create
iscsi objects like targets, disks, clients, etc. The problem is that
once the daemons are started they will accept confifguration requests,
or may want to update the system themself. Those operations can then
conflict with the configure_iscsi.yml tasks that setup objects and we
can end up in crashes due to the kernel being in a unsupported state.
This could also happen during creation, but is less likely due to no
objects being setup yet, so there are no watchers or users accessing the
gws yet. The fix in this patch works for both update and initial setup.
Resolves: https://bugzilla.redhat.com/show_bug.cgi?id=1795806
Signed-off-by: Mike Christie <mchristi@redhat.com>
(cherry picked from commit 77f3b5d51b)
When no monitor group is present in the inventory, this task fails.
This affects only non-containerized deployments.
Signed-off-by: Guillaume Abrioux <gabrioux@redhat.com>
(cherry picked from commit e7bc079405)
This commit allows ceph-ansible 3.2 to be run against ansible 2.7
However, note that running stable-3.2 against ansible 2.7 doesn't get
any testing upstream this might break the playbook, only ansible 2.6 is
officially supported.
Closes: https://bugzilla.redhat.com/show_bug.cgi?id=1781635
Signed-off-by: Guillaume Abrioux <gabrioux@redhat.com>
The [rgw] section in the ceph.conf file or via the ceph_conf_overrides
variable doesn't exist and has no effect.
To apply overrides to all radosgw instances we should use either the
[global] or [client] sections.
Overrides per radosgw instance should still use the
[client.rgw.{instance-name}] section.
Closes: https://bugzilla.redhat.com/show_bug.cgi?id=1794552
Signed-off-by: Dimitri Savineau <dsavinea@redhat.com>
(cherry picked from commit 2f07b85131)
To avoid confusion, let's change the default value from `0.0.0.0` to
`x.x.x.x`.
Users might think setting `0.0.0.0` will make the daemon binding on all
interfaces.
Fixes: #4827
Signed-off-by: Guillaume Abrioux <gabrioux@redhat.com>
(cherry picked from commit fc02fc98eb)
cf8c6a3 moves the 'wait for all osds' task from openstack_config to the
main tasks list.
But the openstack_config code was executed only on the last OSD node.
We don't need to do this check on all OSD node so we need to add set
run_once to true on that task.
Signed-off-by: Dimitri Savineau <dsavinea@redhat.com>
(cherry picked from commit 5bd1cf40eb)
When creating crush rules with device class parameter we need to be sure
that all OSDs are up and running because the device class list is
is populated with this information.
This is now enable for all scenario not openstack_config only.
Signed-off-by: Dimitri Savineau <dsavinea@redhat.com>
(cherry picked from commit cf8c6a3849)
When upgrading from jewel to luminous we can execute the crush rule tasks
only when the 'osd require-osd-release luminous' command.
Signed-off-by: Dimitri Savineau <dsavinea@redhat.com>
This adds device class support to crush rules when using the class key
in the rule dict via the create-replicated sub command.
If the class key isn't specified then we use the create-simple sub
command for backward compatibility.
Closes: https://bugzilla.redhat.com/show_bug.cgi?id=1636508
Signed-off-by: Dimitri Savineau <dsavinea@redhat.com>
(cherry picked from commit ef2cb99f73)
If we want to create crush rules with the create-replicated sub command
and device class then we need to have the OSD created before the crush
rules otherwise the device classes won't exist.
Signed-off-by: Dimitri Savineau <dsavinea@redhat.com>
(cherry picked from commit ed36a11eab)
When ceph_rbd_mirror_configure is set to true we need to ensure that
the required variables aren't empty.
Closes: https://bugzilla.redhat.com/show_bug.cgi?id=1760553
Signed-off-by: Dimitri Savineau <dsavinea@redhat.com>
(cherry picked from commit 4a065cebd7)
We need to explicitly enable the allow_dirfrags flag on cephfs pool
after upgrading to Luminous.
Closes: https://bugzilla.redhat.com/show_bug.cgi?id=1776233
Signed-off-by: Dimitri Savineau <dsavinea@redhat.com>
When using `osd_auto_discovery`, `devices` is built multiple times due
to multiple runs of `ceph-facts` role. It end up with duplicate
instances of a same device in the list.
Using `unique` filter when building the list fixes this issue.
Signed-off-by: Guillaume Abrioux <gabrioux@redhat.com>
(cherry picked from commit 23b1f43897)
This commit adds the support of the ceph-iscsi stable repository when
use ceph_repository community instead of always using the devel
repositories.
We're still using the devel repositories for rtslib and tcmu-runner in
both cases (dev and community).
Signed-off-by: Dimitri Savineau <dsavinea@redhat.com>
The systemd unit script wasn't updated with the new container name
format (without the hostname).
We now have the same start/stop docker commands for all scenarios.
During the device to id OSD migration we need to be sure that the
old container with the hostname are stopped.
Closes: https://bugzilla.redhat.com/show_bug.cgi?id=1780688
Signed-off-by: Dimitri Savineau <dsavinea@redhat.com>
The md devices (RAID software) aren't excluded from the devices list in
the auto discovery scenario.
Closes: https://bugzilla.redhat.com/show_bug.cgi?id=1764601
Signed-off-by: Dimitri Savineau <dsavinea@redhat.com>
(cherry picked from commit 014f51c2a4)
the previous approach was wrong.
checking if `item.key` is in `osd_auto_discovery_exclude` (`['dm-',
'loop']`) is incorrect because it will obviously not match. Therefore,
the condition will return `True` whatever the device we are checking.
Signed-off-by: Guillaume Abrioux <gabrioux@redhat.com>
(cherry picked from commit 8f42007272)
Add a new `osd_auto_discovery_exclude` to give the possibility of
excluding some devices in auto_discovery scenario.
Signed-off-by: Guillaume Abrioux <gabrioux@redhat.com>
(cherry picked from commit 83d7ef777e)
This task used to live in ceph-osd, but we need it defined here to that
ceph-config can use it when trying to determine the number of osds.
Signed-off-by: Andrew Schoen <aschoen@redhat.com>
(cherry picked from commit 88eda479a9)
This will prevent failure of site-docker.yml with configs in doc.
Closes: https://bugzilla.redhat.com/show_bug.cgi?id=1769760
Signed-off-by: VasishtaShastry <vipin.indiasmg@gmail.com>
Co-Authored-By: Guillaume Abrioux <gabrioux@redhat.com>
(cherry picked from commit 9a1f1626c3)
[1] introduced a regression on the fs.aio-max-nr sysctl value condition.
The enable key isn't a boolean but a string because the expression isn't
evaluated.
This string output "(osd_objectstore == 'bluestore')" is always true
because item.enable condition only matches non empty string. So the
sysctl value was applyied for both filestore and bluestore backend.
[2] added the bool filter to the condition but the filter always returns
false on string and the sysctl wasn't applyed at all.
This commit fixes the enable key value by evaluating the value instead
of using the string.
[1] https://github.com/ceph/ceph-ansible/commit/08a2b58
[2] https://github.com/ceph/ceph-ansible/commit/ab54fe2
Signed-off-by: Dimitri Savineau <dsavinea@redhat.com>
(cherry picked from commit ece46d33be)
Guillaume Abrioux
Dimitri Savineau
RPietrzak
ianwatsonrh
John Fulton
Benoît Knecht
Mike Christie
Andrew Schoen
VasishtaShastry
Noah Watkins