Neha Ojha [Thu, 25 Oct 2018 17:45:00 +0000 (17:45 +0000)]
roles: do not limit docker_memory_limit for various daemons
Since we do not have enough data to put valid upper bounds for the memory
usage of these daemons, do not put artificial limits by default. This will
help us avoid failures like OOM kills due to low default values.
Whenever required, these limits can be manually enforced by the user.
More details in
https://bugzilla.redhat.com/show_bug.cgi?id=1638148
since we set `configure_firewall: true` in
`ceph-defaults/defaults/main.yml` there is no need to explicitly set it
in `centos7_cluster` and `docker_cluster` testing scenarios.
Sébastien Han [Thu, 9 Aug 2018 09:32:53 +0000 (11:32 +0200)]
rolling_update: fix upgrade when using fqdn
CLusters that were deployed using 'mon_use_fqdn' have a different unit
name, so during the upgrade this must be used otherwise the upgrade will
fail, looking for a unit that does not exist.
Closes: https://bugzilla.redhat.com/show_bug.cgi?id=1597516 Signed-off-by: Sébastien Han <seb@redhat.com>
Andrew Schoen [Tue, 16 Oct 2018 15:20:54 +0000 (10:20 -0500)]
validate: check the version of python-notario
If the version of python-notario is < 0.0.13 an error message is given
like "TypeError: validate() got an unexpected keyword argument
'defined_keys'", which is not helpful in figuring
out you've got an incorrect version of python-notario.
This check will avoid that situation by telling the user that they need
to upgrade python-notario before they hit that error.
- fix a typo in vagrant_variables that cause a networking issue for
containerized scenario.
- add containerized_deployment: true
- remove a useless block of code: the fact docker_exec_cmd is set in
ceph-defaults which is played right after.
Sébastien Han [Thu, 24 May 2018 17:47:29 +0000 (10:47 -0700)]
infra: add a gather-ceph-logs.yml playbook
Add a gather-ceph-logs.yml which will log onto all the machines from
your inventory and will gather ceph logs. This is not intended to work
on containerized environments since the logs are stored in journald.
Closes: https://bugzilla.redhat.com/show_bug.cgi?id=1582280 Signed-off-by: Sébastien Han <seb@redhat.com>
Sébastien Han [Thu, 27 Sep 2018 14:31:22 +0000 (16:31 +0200)]
infra: rename osd-configure to add-osd and improve it
The playbook has various improvements:
* run ceph-validate role before doing anything
* run ceph-fetch-keys only on the first monitor of the inventory list
* set noup flag so PGs get distributed once all the new OSDs have been
added to the cluster and unset it when they are up and running
Closes: https://bugzilla.redhat.com/show_bug.cgi?id=1624962 Signed-off-by: Sébastien Han <seb@redhat.com>
Sébastien Han [Thu, 27 Sep 2018 14:29:22 +0000 (16:29 +0200)]
ceph-fetch-keys: refact
This commits simplies the usage of the ceph-fetch-keys role. The role
now has a nicer way to find various ceph keys and fetch them on the
ansible server.
Closes: https://bugzilla.redhat.com/show_bug.cgi?id=1624962 Signed-off-by: Sébastien Han <seb@redhat.com>
Andy McCrae [Fri, 5 Oct 2018 13:36:36 +0000 (14:36 +0100)]
Add ability to use a different client container
Currently a throw-away container is built to run ceph client
commands to setup users, pools & auth keys. This utilises
the same base ceph container which has all the ceph services
inside it.
This PR allows the use of a separate container if the deployer
wishes - but defaults to use the same full ceph container.
This can be used for different architectures or distributions,
which may support the the Ceph client, but not Ceph server,
and allows the deployer to build and specify a separate client
container if need be.
Signed-off-by: Andy McCrae <andy.mccrae@gmail.com>
infra: fix wrong condition on firewalld start task
a non skipped task won't have the `skipped` attribute, so `start
firewalld` task will complain about that.
Indeed, `skipped` and `rc` attributes won't exist since the first task
`check firewalld installation on redhat or suse` won't be skipped in
case of non-containerized deployment.
binhong.hua [Wed, 10 Oct 2018 15:24:30 +0000 (23:24 +0800)]
vagrantfile: remove disk path of OSD nodes
osd node's disks will remain on vagrant host,when run "vagrant destroy",
because we use time as a part of disk path, and time on delete not equal time on create.
we already use random_hostname in Libvirt backend,it will create disk
use the hostname as a part of diskname. for example: vagrant_osd2_1539159988_065f15e3e1fa6ceb0770-hda.qcow2.
`ceph_osd_container_stat` might not be set on other osd node.
We must ensure we are on the last node before trying to evaluate
`ceph_osd_container_stat`.
Sébastien Han [Tue, 2 Oct 2018 15:37:06 +0000 (17:37 +0200)]
ceph-handler: change osd container check
Now that the container is named ceph-osd@<id> looking for something that
contains a host is not necessary. This is also backward compatible as it
will continue to match container names with hostname in them.
Sébastien Han [Fri, 28 Sep 2018 16:07:08 +0000 (18:07 +0200)]
osd: ceph-volume activate, just pass the OSD_ID
We don't need to pass the device and discover the OSD ID. We have a
task that gathers all the OSD ID present on that machine, so we simply
re-use them and activate them. This also handles the situation when you
have multiple OSDs running on the same device.
Sébastien Han [Fri, 28 Sep 2018 11:06:18 +0000 (13:06 +0200)]
ceph_volume: add container support for batch command
The batch option got recently added, while rebasing this patch it was
necessary to implement it. So now, the batch option can work on
containerized environments.
Closes: https://bugzilla.redhat.com/show_bug.cgi?id=1630977 Signed-off-by: Sébastien Han <seb@redhat.com>
Andrew Schoen [Tue, 9 Oct 2018 14:04:51 +0000 (10:04 -0400)]
ceph-validate: remove versions checks for bluestore and lvm scenario
These checks will never pass unless ceph_stable_release is passed and
ceph-defaults is run before ceph-validate. Additionally, we don't want
to support deploying jewel upstream at ceph-ansible master.
Fixes: https://bugzilla.redhat.com/show_bug.cgi?id=1637537 Signed-off-by: Andrew Schoen <aschoen@redhat.com>
Andrew Schoen [Mon, 1 Oct 2018 17:51:47 +0000 (12:51 -0500)]
ceph-volume: make the batch action idempotent
The command is run with --report first to see if any OSDs will be
created or not. If they will be, then the command is run. If not, then
changed is set to False and the module exits.
Sébastien Han [Mon, 8 Oct 2018 13:45:58 +0000 (09:45 -0400)]
site: use default value for 'cluster' variable
If someone's cluster name is 'ceph' then the playbook will fail (with no
errors because of ignore_errors) saying it can not find the variable. So
let's declare the default. If the cluster name is different then it'll
be in group_vars and thus there won't be any failre.
Closes: https://bugzilla.redhat.com/show_bug.cgi?id=1636962 Signed-off-by: Sébastien Han <seb@redhat.com>
Sébastien Han [Wed, 3 Oct 2018 11:39:35 +0000 (13:39 +0200)]
switch: copy initial mon keyring
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.
switch: support migration when cluster is scrubbing
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.
'ansible.vars.hostvars.HostVarsVars object' has no attribute u'ansible_interface'
```
the reason is that it will assume `monitor_address_block` isn't defined even on
ceph-mon2 because looking for `monitor_address_block` instead of
`hostvars[host]['monitor_address_block']`, therefore it enters in the condition as default value:
`monitor_interface` is set with default value `'interface'` so the `interface`
variable is built with 'ansible_' + 'interface'. It makes ansible throwing a
confusing message about `'ansible_interface'`.
Allow user to choose between timesyncd, chronyd and ntpd
Installation will default to timesyncd since it is distributed as
part of the systemd installation for most distros.
Added note indicating NTP daemon type is not used for containerized
deployments.
Fixes issue #3086 on Github
Signed-off-by: Benjamin Cherian <benjamin_cherian@amat.com>
Mike Christie [Fri, 28 Sep 2018 21:23:10 +0000 (16:23 -0500)]
igw: valid client CHAP settings.
The linux kernel target layer, LIO, does not support the iscsi target to
mix ACLs that have chap enabled and disabled under the same tpg. This
patch adds a check and fails if this type of setup is detected.
This fixes Red Hat BZ:
https://bugzilla.redhat.com/show_bug.cgi?id=1615088
Signed-off-by: Mike Christie <mchristi@redhat.com>
projects / ceph-ansible.git / log