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>
Sébastien Han [Fri, 27 Jul 2018 14:56:09 +0000 (16:56 +0200)]
add ceph-handler role
The role contains all the handlers for Ceph services. We decided to
leave ceph-defaults role with variables and a few facts only. This is
useful when organizing the site.yml files and also adding the known
variables to infrastructure-playbooks.
Sébastien Han [Thu, 27 Sep 2018 07:57:26 +0000 (09:57 +0200)]
default: use bluestore as default object store
All tooling in Ceph is defaulting to use the bluestore objectstore for provisioning OSDs, there is no good reason for ceph-ansible to continue to default to filestore.
Closes: https://github.com/ceph/ceph-ansible/issues/3149 Closes: https://bugzilla.redhat.com/show_bug.cgi?id=1633508 Signed-off-by: Sébastien Han <seb@redhat.com>
This commit brings back this condition which is essential to ensure we
won't hit a false positive result in the `when` condition for the check
PGs task.
upgrade: consider all 'active+clean' states as valid pgs
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 before a
rolling update which is a problem because it pauses an important data
integrity operation until the end of the rolling upgrade.
This commit allows an upgrade even while PGs are scrubbing.
tests: add an RGW node on osd0 for ooo-collocation
get more coverage by adding an RGW daemon collocated on osd0.
We've missed a bug in the past which could have been caught earlier in
the CI.
Let's add this additional daemon in order to have a better coverage.
Matthew Vernon [Wed, 19 Sep 2018 12:26:26 +0000 (13:26 +0100)]
restart_osd_daemon.sh.j2 - consider active+clean+* pgs as OK
After restarting each OSD, restart_osd_daemon.sh checks that the
cluster is in a good state before moving on to the next one. One of
the checks it does is that the number of pgs in the state
"active+clean" is equal to the total number of pgs in the cluster.
On large clusters (e.g. we have 173,696 pgs), it is likely that at
least one pg will be scrubbing and/or deep-scrubbing at any one
time. These pgs are in state "active+clean+scrubbing" or
"active+clean+scrubbing+deep", so the script was erroneously not
including them in the "good" count. Similar concerns apply to
"active+clean+snaptrim" and "active+clean+snaptrim_wait".
Fix this by considering as good any pg whose state contains
active+clean. Do this as an integer comparison to num_pgs in pgmap.
(could this be backported to at least stable-3.0 please?)
Closes: #2008 Signed-off-by: Matthew Vernon <mv3@sanger.ac.uk>
Matthew Vernon [Fri, 21 Sep 2018 16:55:01 +0000 (17:55 +0100)]
restart_osd_daemon.sh.j2 - Reset RETRIES between calls of check_pgs
Previously RETRIES was set (by default to 40) once at the start of the
script; this meant that it would only ever wait for up to 40 lots of
30s across *all* the OSDs on a host before bombing out. In fact, we
want to be prepared to wait for the same amount of time after each OSD
restart for the clusters' pgs to be happy again before continuing.
Closes: #3154 Signed-off-by: Matthew Vernon <mv3@sanger.ac.uk>
In case of an OpenStack "box", the Vagrantfile intend to check the
existence of os_networks and os_floating_ip_pool settings in
vagrant_variables.yml and pass them to the provider if they are set.
Due to two typos in the Vagrantfile this is not working as it checks the
wrong variable names.
This commit fixes the typos so these settings can be used.
Andrew Schoen [Tue, 18 Sep 2018 20:12:59 +0000 (15:12 -0500)]
ceph-config: calculate num_osds for the lvm batch scenario
For now our best guess is to count the number of devices and multiply
by osds_per_device. Ideally we'd like to run ceph-volume lvm batch
--report and get the number of OSDs that way, but currently we need
a ceph.conf in place already before we can do that. There is a tracker
ticket that would allow os to get around the need for a ceph.conf:
http://tracker.ceph.com/issues/36088
Fixes: https://github.com/ceph/ceph-ansible/issues/3135 Signed-off-by: Andrew Schoen <aschoen@redhat.com>
Andrew Schoen [Tue, 18 Sep 2018 14:36:08 +0000 (09:36 -0500)]
ceph-config: default devices and lvm_volumes when setting num_osds
This avoids errors when the osd scenario choosen does not require
setting devices or lvm_volumes. The default values for these are not
set because they exist in the ceph-osd role, not ceph-defaults.
BlueStore's cache is sized conservatively by default, so that it does
not overwhelm under-provisioned servers. The default is 1G for HDD, and
3G for SSD.
To replace the page cache, as much memory as possible should be given to
BlueStore. This is required for good performance. Since ceph-ansible
knows how much memory a host has, it can set
`bluestore cache size = max(total host memory / num OSDs on this host * safety
factor, 1G)`
Due to fragmentation and other memory use not included in bluestore's
cache, a safety factor of 0.5 for dedicated nodes and 0.2 for
hyperconverged nodes is recommended.
- Adds loop in bash to satisfy the 1:n relation between `osd_hosts` and the
different device lists.
- Fixes some container name which were using the host hostname instead
of the actual container one.
installs the cli package but does not start and enable the
rbd-target-api daemon needed for gwcli to communicate with the igw
nodes. This patch just enables and starts it for the non-container
setup. The container setup is already doing this.
This fixes bz https://bugzilla.redhat.com/show_bug.cgi?id=1613963
Signed-off-by: Mike Christie <mchristi@redhat.com>
shrink-osd: fix purge osd on containerized deployment
ce1dd8d introduced the purge osd on containers but it was incorrect.
`resolve parent device` and `zap ceph osd disks` tasks must be delegated to
their respective OSD nodes.
Indeed, they were run on the ansible node, it means it was trying to
resolve parent devices from this node where it should be done on OSD
nodes.
```
Exception occurred:
File
"/home/jenkins-build/build/workspace/ceph-ansible-docs-pull-requests/docs/.tox/docs/lib/python2.7/site-packages/sphinx/highlighting.py", line 26, in <module>
from sphinx.ext import doctest
SyntaxError: unqualified exec is not allowed in function 'run' it contains a nested function with free variables (doctest.py, line 97)
```
Andrew Schoen [Thu, 6 Sep 2018 19:00:56 +0000 (14:00 -0500)]
ceph_volume: adds the osds_per_device parameter
If this is set to anything other than the default value of 1 then the
--osds-per-device flag will be used by the batch command to define how
many osds will be created per device.
mon: fix `ExecStartPre` option in systemd unit file
This command line is not supported.
According to official documentation:
```
Note that shell command lines are not directly supported.
If shell command lines are to be used,
they need to be passed explicitly to a shell implementation of some kind.
```
Tom Barron [Sat, 1 Sep 2018 14:32:51 +0000 (10:32 -0400)]
run rados cmd in container if containerized deployment
When ceph-nfs is deployed containerized and ceph-common is not
installed on the host the start_nfs task fails because the rados
command is missing on the host.
Run rados commands from a ceph container instead so that
they will succeed.
Markos Chandras [Wed, 29 Aug 2018 10:56:16 +0000 (11:56 +0100)]
roles: ceph-rgw: Enable the ceph-radosgw target
If the ceph-radosgw target is not enabled, then enabling the
ceph-radosgw@ service has no effect since nothing will pull
it on the next reboot. As such, we need to ensure that the
target is enabled.
Andy McCrae [Thu, 30 Aug 2018 07:53:36 +0000 (08:53 +0100)]
Dont run client dummy container on non-x86_64 hosts
The dummy client container currently wont work on non-x86_64 hosts.
This PR creates a filtered client group that contains only hosts
that are x86_64 - which can then be the group to run the
dummy container against.
This is for the specific case of a containerized_deployment where
there is a mixture of non-x86_64 hosts and x86_64 hosts. As such
the filtered group will contain all hosts when running with
containerized_deployment: false.
Currently ppc64le is not supported for Ceph server components.
Signed-off-by: Andy McCrae <andy.mccrae@gmail.com>
Sébastien Han [Mon, 27 Aug 2018 17:20:32 +0000 (10:20 -0700)]
sites: fix conditonnal
Same problem again... ceph_release_num[ceph_release] is only set in
ceph-docker-common/common roles so putting the condition on that role
will never work. Removing the condition.
The downside of this is we will be installing packages and then skip the
role on the node.
Closes: https://bugzilla.redhat.com/show_bug.cgi?id=1622210 Signed-off-by: Sébastien Han <seb@redhat.com>
The conditional check 'ceph_current_status.rc == 0' failed. The error was: error while evaluating conditional (ceph_current_status.rc == 0): 'dict object' has no attribute 'rc'
The error appears to have been in '/home/jenkins-build/build/workspace/ceph-ansible-nightly-luminous-stable-3.1-ooo_collocation/roles/ceph-defaults/tasks/facts.yml': line 74, column 3, but may
be elsewhere in the file depending on the exact syntax problem.
The offending line appears to be:
- name: set_fact ceph_current_status (convert to json)
^ here
```
From https://2.jenkins.ceph.com/view/ceph-ansible-stable3.1/job/ceph-ansible-nightly-luminous-stable-3.1-ooo_collocation/37/consoleFull#1765217701b5dd38fa-a56e-4233-a5ca-584604e56e3a
What's happening here is all the hosts excepts the clients are running atomic, so here: https://github.com/ceph/ceph-ansible/blob/master/site-docker.yml.sample#L62
The condition will skipped all the nodes excepts the clients, thus when running ceph-default, the task "is ceph running already?" is skipped but the task above needs the rc of the skipped task.
This is not an error from the playbook, it's a CI setup issue.
Sébastien Han [Tue, 21 Aug 2018 18:50:31 +0000 (20:50 +0200)]
defaults: fix rgw_hostname
A couple if things were wrong in the initial commit:
* ceph_release_num[ceph_release] >= ceph_release_num['luminous'] will
never work since the ceph_release fact is set in the roles after. So
either ceph-common or ceph-docker-common set it
* we can easily re-use the initial command to check if a cluster is
running, it's more elegant than running it twice.
* set the fact rgw_hostname on rgw nodes only
Closes: https://bugzilla.redhat.com/show_bug.cgi?id=1618678 Signed-off-by: Sébastien Han <seb@redhat.com>
Sébastien Han [Tue, 21 Aug 2018 09:15:44 +0000 (11:15 +0200)]
rolling_upgrade: set sortbitwise properly
Running 'osd set sortbitwise' when we detect a version 12 of Ceph is
wrong. When OSD are getting updated, even though the package is updated
they won't send their updated version (12) and will stick with 10 if the
command is not applied. So we have to check if OSD are sending a version
10 and then run the command to unlock the OSDs.
Closes: https://bugzilla.redhat.com/show_bug.cgi?id=1600943 Signed-off-by: Sébastien Han <seb@redhat.com>
Sébastien Han [Mon, 20 Aug 2018 13:53:03 +0000 (15:53 +0200)]
iscsi group name preserve backward compatibility
Recently we renamed the group_name for iscsi iscsigws where previously
it was named iscsi-gws. Existing deployments with a host file section
with iscsi-gws must continue to work.
This commit adds the old group name as a backoward compatility, no error
from Ansible should be expected, if the hostgroup is not found nothing
is played.
Close: https://bugzilla.redhat.com/show_bug.cgi?id=1619167 Signed-off-by: Sébastien Han <seb@redhat.com>
Sébastien Han [Mon, 20 Aug 2018 12:41:06 +0000 (14:41 +0200)]
take-over-existing-cluster: do not call var_files
We were using var_files long ago when default variables were not in
ceph-defaults, now the role exists this is not need. Moreover having
these two var files added:
Markos Chandras [Wed, 15 Aug 2018 06:55:49 +0000 (09:55 +0300)]
roles: ceph-defaults: Check if 'rgw' attribute exists for rgw_hostname
If there are no services on the cluster, then the 'rgw' could be missing
and the task is failing with the following problem:
msg": "The task includes an option with an undefined variable.
The error was: 'dict object' has no attribute 'rgw'
We fix this by checking the existence of the 'rgw' attribute. If it's
missing, we skip the task since the role already contains code to set
a good default rgw_hostname.
Markos Chandras [Tue, 14 Aug 2018 06:52:04 +0000 (09:52 +0300)]
roles: ceph-defaults: Delegate cluster information task to monitor node
Since commit f422efb1d6b56ce56a7d39a21736a471e4ed357 ("config: ensure
rgw section has the correct name") we observe the following failures in
new Ceph deployment with OpenStack-Ansible
fatal: [aio1_ceph-rgw_container-fc588f0a]: FAILED! => {"changed": false,
"cmd": "ceph --cluster ceph -s -f json", "msg": "[Errno 2] No such file
or directory"
This is because the task executes 'ceph' but at this point no package
installation has happened. Packages are normally installed in the
'ceph-common' role which runs after the 'ceph-defaults' one.
Since we are looking to obtain cluster information, the task should be
delegated to a monitor node similar to other tasks in that role
Andrew Schoen [Wed, 8 Aug 2018 22:12:30 +0000 (17:12 -0500)]
tests: adds a testing scenario for lv-create and lv-teardown
Using an explicitly named testing environment name allows us to have a
specific [testenv] block for this test. This greatly simplifies how it will
work as it doesn't really anything from the ceph cluster tests.
Andrew Schoen [Wed, 8 Aug 2018 21:43:55 +0000 (16:43 -0500)]
lv-create: use the template module to write log file
The copy module will not expand the template and render the variables
included, so we must use template.
Creating a temp file and using it locally means that you must run the
playbook with sudo privledges, which I don't think we want to require.
This introduces a logfile_path variable that the user can use to control
where the logfile is written to, defaulting to the cwd.
Sébastien Han [Mon, 13 Aug 2018 13:59:25 +0000 (15:59 +0200)]
rolling_update: register container osd units
Before running the upgrade, let's call systemd to collect unit names
instead of relaying on the device list. This is more accurate and fix
the osd_auto_discovery scenario too.
Closes: https://bugzilla.redhat.com/show_bug.cgi?id=1613626 Signed-off-by: Sébastien Han <seb@redhat.com>
This commit was giving a new failure later during the rolling_update
process. Basically, this was modifying the list of devices and started
impacting the ceph-osd itself. The modification to accomodate the
osd_auto_discovery parameter should happen outside of the ceph-osd.
Also we are trying to not play ceph-osd role during the rolling_update
process so we can speed up the upgrade.
fqdn configuration possibility caused a lot of trouble, it's adding a
lot of complexity because of multiple cases and the relation between
ceph-ansible and ceph-container. Moreover, there is no benefit for such
a feature.
the ceph.conf.j2 always assumes the hostname used to register the
radosgw in the servicemap is equivalent to `{{ ansible_hostname }}`
which returns the shortname form.
We need to detect which form of the hostname was used in case of already
deployed cluster and update the ceph.conf accordingly.
Sébastien Han [Fri, 10 Aug 2018 09:08:14 +0000 (11:08 +0200)]
mon: fix calamari initialisation
If calamari is already installed and ceph has been upgraded to a higher
version the initialisation will fail later. So if we detect the
calamari-server is too old compare to ceph_rhcs_version we try to update
it.
Closes: https://bugzilla.redhat.com/show_bug.cgi?id=1601755 Signed-off-by: Sébastien Han <seb@redhat.com>
Andrew Schoen [Thu, 9 Aug 2018 15:40:16 +0000 (10:40 -0500)]
lvm: fix condition when selecting which scenario to run
devices and lvm_volumes will always be defined, so we need to instead
check it's length before deciding to run the scenario.
This fixes the failure here:
https://2.jenkins.ceph.com/job/ceph-ansible-prs-luminous-bluestore_lvm_osds/86/consoleFull#1667273050b5dd38fa-a56e-4233-a5ca-584604e56e3a
Sébastien Han [Thu, 9 Aug 2018 13:18:34 +0000 (15:18 +0200)]
osd: generate device list for osd_auto_discovery on rolling_update
rolling_update relies on the list of devices when performing the restart
of the OSDs. The task that is builind the devices list out of the
ansible_devices dict only runs when there are no partitions on the
drives. However during an upgrade the OSD are already configured, they
have been prepared and have partitions so this task won't run and thus
the devices list will be empty, skipping the restart during
rolling_update. We now run the same task under different requirements
when rolling_update is true and build a list when:
* osd_auto_discovery is true
* rolling_update is true
* ansible_devices exists
* no dm/lv are part of the discovery
* the device is not removable
* the device has more than 1 sector
Closes: https://bugzilla.redhat.com/show_bug.cgi?id=1613626 Signed-off-by: Sébastien Han <seb@redhat.com>
Andrew Schoen [Mon, 6 Aug 2018 20:14:53 +0000 (15:14 -0500)]
ceph-osd: adds crush_device_class config option
This is used with the lvm osd scenario. When using devices you need the
option to set the crush device class for all of the OSDs that are
created from those devices.
Andrew Schoen [Fri, 3 Aug 2018 16:15:58 +0000 (11:15 -0500)]
ceph-volume: implement the 'lvm batch' subcommand
This adds the action 'batch' to the ceph-volume module so that we can
run the new 'ceph-volume lvm batch' subcommand. A functional test is
also included.
If devices is defind and osd_scenario is lvm then the 'ceph-volume lvm
batch' command will be used to create the OSDs.
projects / ceph-ansible.git / log