This commit adds the pg autoscaler support.
The structure for pool definition has now two additional attributes
`pg_autoscale_mode` and `target_size_ratio`, eg:
```
test:
name: "test"
pg_num: "{{ osd_pool_default_pg_num }}"
pgp_num: "{{ osd_pool_default_pg_num }}"
rule_name: "replicated_rule"
application: "rbd"
type: 1
erasure_profile: ""
expected_num_objects: ""
size: "{{ osd_pool_default_size }}"
min_size: "{{ osd_pool_default_min_size }}"
pg_autoscale_mode: False
target_size_ratio": 0.1
```
when `pg_autoscale_mode` is `True` user has to set a decent value in
`target_size_ratio`.
Given that it's a new feature, it's still disabled by default.
Closes: https://bugzilla.redhat.com/show_bug.cgi?id=1782253
Signed-off-by: Guillaume Abrioux <gabrioux@redhat.com>
Currently, the command executed is wrong, eg:
```
cmd:
- podman
- exec
- ceph-mon-controller-0
- ceph
- --cluster
- ceph
- osd
- pool
- create
- volumes
- '32'
- '32'
- replicated_rule
- '1'
delta: '0:00:01.625525'
end: '2020-02-27 16:41:05.232705'
item:
```
From documentation, the osd pool creation command is :
```
ceph osd pool create {pool-name} {pg-num} [{pgp-num}] [replicated] \
[crush-rule-name] [expected-num-objects]
ceph osd pool create {pool-name} {pg-num} {pgp-num} erasure \
[erasure-code-profile] [crush-rule-name] [expected_num_objects]
```
it means we pass '1' (from item.type) as value for
`expected_num_objects` by default which is very likely not what we want.
Also, this commit modifies the default value when no `rule_name` is set
to use the existing variable `osd_pool_default_crush_rule`
Closes: https://bugzilla.redhat.com/show_bug.cgi?id=1808495
Signed-off-by: Guillaume Abrioux <gabrioux@redhat.com>
Looks like we are still seeing issue [1].
Let's increase this value to unlock the CI (however, it still needs to
be investigated).
Typical error (see [1] for further details) :
```
[root@osd2 ~]# ceph-volume --cluster ceph lvm batch --filestore --yes --journal-size '2048' /dev/sda /dev/sdb --journal-devices /dev/sdc
Running command: /sbin/vgcreate --force --yes ceph-journals-817ef90b-77ac-4f52-b8a9-30893849fb78 /dev/sdc
stdout: Physical volume "/dev/sdc" successfully created.
stdout: Volume group "ceph-journals-817ef90b-77ac-4f52-b8a9-30893849fb78" successfully created
--> Refusing to continue with configured size for journal
--> RuntimeError: journal sizes must be larger than 2GB, detected: 1024.00 MB
```
[1] https://tracker.ceph.com/issues/41374
Signed-off-by: Guillaume Abrioux <gabrioux@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 shrink scenarios don't need the docker variables (except for OSD).
Removing pytest for shrink-mgr.
Adding environment variables for xxx_to_kill ansible variable.
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>
If the user provides manually the key value for a specific keyring then
there's not valation on the content which could lead to unexpected
failures in the ceph_key module.
Closes: #5104
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 using the firewalld ansible module we need to be sure that the
python bindings are installed.
Signed-off-by: Dimitri Savineau <dsavinea@redhat.com>
Since ansible 2.9 the firewalld task could not be used with service and
source in the same time anymore.
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>
This change introduces a new set of tasks to configure the
ceph dashboard backend and listen just on the mgr related
subnet (and not on '*'). For the same reason the proper
server address is added in both prometheus and alertmanger
systemd units.
This patch also adds the "dashboard_frontend_vip" parameter
to make sure we're able to support the HA model when multiple
grafana instances are deployed.
Closes: https://bugzilla.redhat.com/show_bug.cgi?id=1792230
Signed-off-by: Francesco Pantano <fpantano@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>
Since [1] if a rgw user already exists then the radosgw-admin user create
command will return an error instead of modifying the current user.
We were already doing separated tasks for create and get operation but
only for multisite configuration but it's not enough.
Instead we should do the get task first and depending on the result
execute the create.
This commit also adds missing run_once and delegate_to statement.
[1] https://github.com/ceph/ceph/commit/269e9b9
Signed-off-by: Dimitri Savineau <dsavinea@redhat.com>
Allow SSL certificate & key contents to be written to the path
specified by radosgw_frontend_ssl_certificate. This permits a
certificate to be deployed & renewal of expired certificates
through ceph-ansible.
Signed-off-by: Sam Choraria <sam.choraria@bbc.co.uk>
The ad7a5da commit introduced a regression when using TLS on haproxy
via the haproxy_frontend_ssl_certificate variable.
This cause the "stats socket" and the "tune.ssl.default-dh-param"
parameters to be on the same line resulting haproxy failing to start.
[ALERT] 351/140240 (21388) : parsing [xxxxx] : 'stats socket' : unknown
keyword 'tune.ssl.default-dh-param'. Registered
[ALERT] 351/140240 (21388) : Fatal errors found in configuration.
Fixes: #4869
Signed-off-by: Florian Faltermeier <florian.faltermeier@uibk.ac.at>
The rgw user creation for the Ceph dashboard integration shouldn't be
created on secondary rgw zones.
Closes: #4707
Closes: https://bugzilla.redhat.com/show_bug.cgi?id=1794351
Signed-off-by: Dimitri Savineau <dsavinea@redhat.com>
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>
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>
Using ceph_dev_branch and ceph_dev_sha1 for configuring ceph-iscsi
repositories from shaman doesn't make sense because the ceph devel
branches and sha1 aren't compatible with ceph-iscsi devel.
Instead we could rely on the master branch and the latest sha1.
Currently it's not possible to using a custom ceph branch/sha1 value
with iscsi setup otherwise the repository setup will fail.
Signed-off-by: Dimitri Savineau <dsavinea@redhat.com>
The ceph_nfs_ceph_user variable is a string for the ceph-nfs role but a
list in ceph-client role.
6a6785b introduced a confusion between both variable type in the ceph-nfs
role for external ceph with ganesha.
Closes: https://bugzilla.redhat.com/show_bug.cgi?id=1801319
Signed-off-by: Dimitri Savineau <dsavinea@redhat.com>
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>
The grafana_{crt,key} aren't boolean variables but strings. The default
value is an empty string so we should do the conditional on the string
length instead of the bool filter
Closes: #5053
Signed-off-by: Dimitri Savineau <dsavinea@redhat.com>
Because we are relying on docker|podman for managing containers then we
don't need systemd to manage the process (like kill).
Signed-off-by: Dimitri Savineau <dsavinea@redhat.com>
When using multiple grafana hosts then we push set the grafana and
prometheus URL and push the dashboard layout to a single node.
grafana_server_addrs is the list of all grafana nodes and used during
the ceph-dashboard role (on mgr/mon nodes).
grafana_server_addr is the current grafana node used during the
ceph-grafana and ceph-prometheus role (on grafana-server nodes).
We don't have the grafana_server_addr fact duplication code between
external vs collocated nodes.
Closes: https://bugzilla.redhat.com/show_bug.cgi?id=1784011
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>
This wasn't necesarry. The container image was fixed on the
RedHat's registry
This reverts commit 3bd250c742.
Signed-off-by: Dimitri Savineau <dsavinea@redhat.com>
This was used before the CentOS 8 requirement when using CentOS 7
atomic which has both docker and podman installed.
Signed-off-by: Dimitri Savineau <dsavinea@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
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>
Guillaume Abrioux
Dimitri Savineau
Francesco Pantano
Benoît Knecht
Sam Choraria
Ali Maredia
Florian Faltermeier
John Fulton
wujie1993
Mike Christie