]> git-server-git.apps.pok.os.sepia.ceph.com Git - ceph-ansible.git/log
ceph-ansible.git
8 years agoceph-common: improve error message on restart osd daemon script 1523/head
Alfredo Deza [Fri, 12 May 2017 15:01:03 +0000 (11:01 -0400)]
ceph-common: improve error message on restart osd daemon script

Signed-off-by: Alfredo Deza <adeza@redhat.com>
(cherry picked from commit b04d18988e7f4c3da332af8941550ba0af18365a)

8 years agoMerge pull request #1506 from ceph/backport-1505
Sébastien Han [Wed, 10 May 2017 16:27:51 +0000 (18:27 +0200)]
Merge pull request #1506 from ceph/backport-1505

backport: "remove the install-ansible.sh script #1505"

8 years agoremove the install-ansible.sh script 1506/head
Andrew Schoen [Wed, 10 May 2017 16:16:47 +0000 (11:16 -0500)]
remove the install-ansible.sh script

I'm unsure why this is needed, users should be able to install ansible
on their own in the method they prefer.

Signed-off-by: Andrew Schoen <aschoen@redhat.com>
(cherry picked from commit 1dc0fab21ca3126bc06aa54af4daae87feedeeee)

8 years agoMerge pull request #1491 from ceph/backport-1490 v2.2.4
Andrew Schoen [Wed, 3 May 2017 01:24:06 +0000 (20:24 -0500)]
Merge pull request #1491 from ceph/backport-1490

backport: "tests: allow for insecure docker registries when testing rhcs #1490"

8 years agotests: allow for insecure docker registries when testing rhcs 1491/head
Andrew Schoen [Tue, 2 May 2017 20:49:50 +0000 (15:49 -0500)]
tests: allow for insecure docker registries when testing rhcs

Signed-off-by: Andrew Schoen <aschoen@redhat.com>
(cherry picked from commit 2fada9bd6b0b0099f191be875edeee23a2fe3daa)

Conflicts:
tests/functional/rhcs_setup.yml

8 years agoMerge pull request #1489 from ceph/backport-1487
Andrew Schoen [Wed, 3 May 2017 01:16:01 +0000 (20:16 -0500)]
Merge pull request #1489 from ceph/backport-1487

Backport: "tests: actually test jewel when we say we are #1487"

8 years agotests: switch to using key/value extra-vars instead of json 1489/head
Andrew Schoen [Tue, 2 May 2017 16:15:23 +0000 (11:15 -0500)]
tests: switch to using key/value extra-vars instead of json

I continue to have issues with extra-vars as json. The latest issue
being that the ceph_docker_image_tag config option included in the json
was being ignored. I can't find the root cause, by using the key/value
format seems to work.

I've also removed several options here to simply the interface. We can
add those back if they become necessary.

Signed-off-by: Andrew Schoen <aschoen@redhat.com>
(cherry picked from commit e384b2498755e5537595158ca498e03013e33673)

8 years agotests: actually test a jewel container for the jewel-* docker tests
Andrew Schoen [Mon, 1 May 2017 19:08:55 +0000 (14:08 -0500)]
tests: actually test a jewel container for the jewel-* docker tests

Signed-off-by: Andrew Schoen <aschoen@redhat.com>
(cherry picked from commit 1fcc3283d8891055ae394d03f3ded1f6a4c8ec64)

8 years agoMerge pull request #1488 from ceph/backport-1427
Andrew Schoen [Mon, 1 May 2017 20:29:06 +0000 (15:29 -0500)]
Merge pull request #1488 from ceph/backport-1427

tests: enable the rhcs-7-extras-nightly repo on test nodes

8 years agotests: enable the rhcs-7-extras-nightly repo on test nodes 1488/head
Andrew Schoen [Mon, 10 Apr 2017 18:46:03 +0000 (13:46 -0500)]
tests: enable the rhcs-7-extras-nightly repo on test nodes

Signed-off-by: Andrew Schoen <aschoen@redhat.com>
(cherry picked from commit 35aab1274a5ec6ad696040965c87a87310bb019e)

8 years agoMerge pull request #1486 from ceph/backport-1485
Andrew Schoen [Mon, 1 May 2017 20:17:37 +0000 (15:17 -0500)]
Merge pull request #1486 from ceph/backport-1485

Backport: "tests: test all docker osds are up and in #1485"

8 years agotests: test that all docker OSDs are up and in 1486/head
Andrew Schoen [Mon, 1 May 2017 15:45:44 +0000 (10:45 -0500)]
tests: test that all docker OSDs are up and in

Signed-off-by: Andrew Schoen <aschoen@redhat.com>
(cherry picked from commit 10d39e98dde5042d98d234887ead17ff7d278e55)

8 years agotests: adds a way to mark docker only tests
Andrew Schoen [Mon, 1 May 2017 15:45:03 +0000 (10:45 -0500)]
tests: adds a way to mark docker only tests

Signed-off-by: Andrew Schoen <aschoen@redhat.com>
(cherry picked from commit 06b36fddf9689659b97c480b8b6644775fac907b)

8 years agoMerge pull request #1478 from ceph/backport_1477 v2.2.3
Andrew Schoen [Thu, 27 Apr 2017 15:32:26 +0000 (10:32 -0500)]
Merge pull request #1478 from ceph/backport_1477

Backport: Common: Fix condition on osd restart handler

8 years agoCommon: Fix condition on osd restart handler 1478/head
Guillaume Abrioux [Thu, 27 Apr 2017 14:24:21 +0000 (16:24 +0200)]
Common: Fix condition on osd restart handler

Fix the condition on the osd restart handler.

Signed-off-by: Guillaume Abrioux <gabrioux@redhat.com>
8 years agoMerge pull request #1474 from ceph/wip-2.2-rgw-tuning v2.2.2
Andrew Schoen [Tue, 25 Apr 2017 22:12:52 +0000 (17:12 -0500)]
Merge pull request #1474 from ceph/wip-2.2-rgw-tuning

rgw: set tuning parameters

8 years agorgw: set tuning parameters 1474/head
Ali Maredia [Wed, 19 Apr 2017 16:30:40 +0000 (12:30 -0400)]
rgw: set tuning parameters

Change civetweb_num_thread default to 100

Add capability to override number of pgs for
rgw pools.

Add ceph.conf vars to enable default bucket
object quota at users choosing into the ceph.conf.j2
template

Resolves: rhbz#1437173
Resolves: rhbz#1391500

Signed-off-by: Ali Maredia <amaredia@redhat.com>
8 years agoMerge pull request #1473 from ceph/backport-1464
Andrew Schoen [Tue, 25 Apr 2017 19:24:12 +0000 (14:24 -0500)]
Merge pull request #1473 from ceph/backport-1464

backport: "purge-cluster: remove unneeded tasks #1464"

8 years agopurge-cluster: check for any running ceph process after purge 1473/head
Andrew Schoen [Tue, 25 Apr 2017 12:35:13 +0000 (07:35 -0500)]
purge-cluster: check for any running ceph process after purge

Signed-off-by: Andrew Schoen <aschoen@redhat.com>
(cherry picked from commit 5a3f95dfc1bc9bba7ad62999969997ce0a27da13)

8 years agopurge-cluster: we don't support sysv or upstart anymore
Andrew Schoen [Fri, 21 Apr 2017 22:14:38 +0000 (15:14 -0700)]
purge-cluster: we don't support sysv or upstart anymore

Now that ceph-ansible only supports > jewel we don't need
to bother with sysv or upstart

Signed-off-by: Andrew Schoen <aschoen@redhat.com>
(cherry picked from commit 26bdd59f5d445db7991eecd0d13f7c8615bc4812)

8 years agopurge-cluster: do not need to check for running ceph processes
Andrew Schoen [Fri, 21 Apr 2017 22:12:46 +0000 (15:12 -0700)]
purge-cluster: do not need to check for running ceph processes

Signed-off-by: Andrew Schoen <aschoen@redhat.com>
(cherry picked from commit 7ca2bddcce76dd505b169457fa8727b62b861756)

8 years agopurge-cluster: no need to remove ceph.target
Andrew Schoen [Fri, 21 Apr 2017 22:11:03 +0000 (15:11 -0700)]
purge-cluster: no need to remove ceph.target

The package uninstalls will stop ceph.target

Signed-off-by: Andrew Schoen <aschoen@redhat.com>
(cherry picked from commit aac79df3b376afcb70b4e8888a72ad016079576b)

8 years agoMerge pull request #1470 from ceph/backport-1466
Andrew Schoen [Tue, 25 Apr 2017 18:03:30 +0000 (13:03 -0500)]
Merge pull request #1470 from ceph/backport-1466

backport-1466: mon: upgrade calamari when running rolling_update playbook

8 years agomon: upgrade calamari when running rolling_update playbook 1470/head
Sébastien Han [Mon, 24 Apr 2017 09:04:06 +0000 (11:04 +0200)]
mon: upgrade calamari when running rolling_update playbook

Prior to this change, ansible was only checking for the existence of the
package, now if upgrade_ceph_packages is true this means we are
performing an upgrade.

Fixes: https://bugzilla.redhat.com/show_bug.cgi?id=1442016
Signed-off-by: Sébastien Han <seb@redhat.com>
(cherry picked from commit 84d96be197617ff8cdf8141e4a52657f172b2809)

8 years agoMerge pull request #1467 from ceph/backport_restart_daemons
Andrew Schoen [Tue, 25 Apr 2017 12:27:15 +0000 (07:27 -0500)]
Merge pull request #1467 from ceph/backport_restart_daemons

Common: Restore check_socket

8 years agoCommon: Restore check_socket 1467/head
Guillaume Abrioux [Tue, 18 Apr 2017 10:40:43 +0000 (12:40 +0200)]
Common: Restore check_socket

Restore the check_socket that was removed by `5bec62b`.
This commit also improves the logging in `restart_*_daemon.sh` scripts

Signed-off-by: Guillaume Abrioux <gabrioux@redhat.com>
8 years agoMerge pull request #1450 from ceph/backport-1439 v2.2.1
Andrew Schoen [Mon, 17 Apr 2017 14:38:55 +0000 (09:38 -0500)]
Merge pull request #1450 from ceph/backport-1439

Backport: 'tests: fix ceph_rhcs setup #1439'

8 years agoMerge pull request #1451 from ceph/doc-static-dirs
Andrew Schoen [Thu, 13 Apr 2017 20:25:47 +0000 (15:25 -0500)]
Merge pull request #1451 from ceph/doc-static-dirs

Doc static dirs

8 years agodocs: force an empty _templates dir so sphinx is happy 1451/head
Alfredo Deza [Tue, 11 Apr 2017 13:08:45 +0000 (09:08 -0400)]
docs: force an empty _templates dir so sphinx is happy

Signed-off-by: Alfredo Deza <adeza@redhat.com>
(cherry picked from commit 66a9817e56d69b046425dc04a03869fb9a9d8369)
(cherry picked from commit df315fef897705c77bf12780d31ddf17376388fb)

8 years agodocs: force an empty _static dir so sphinx is happy
Alfredo Deza [Tue, 11 Apr 2017 13:08:08 +0000 (09:08 -0400)]
docs: force an empty _static dir so sphinx is happy

Signed-off-by: Alfredo Deza <adeza@redhat.com>
(cherry picked from commit aacbbf0efe802e9282793dfc8bf7b175596c5aad)
(cherry picked from commit d26ed99dcc1b930a3001ca1bc71f034b9da4f091)

8 years agotests: print contents of group_vars/all after modification 1449/head 1450/head
Andrew Schoen [Wed, 12 Apr 2017 14:23:21 +0000 (09:23 -0500)]
tests: print contents of group_vars/all after modification

This is just nice to see in the test output so we know exactly what
configuration is going to be used.

Signed-off-by: Andrew Schoen <aschoen@redhat.com>
(cherry picked from commit 0e6d89b9db1b4677db42bf04c88547ebc00e1194)

8 years agoceph-common: is ceph_rchs is True do not include install_on_redhat.yml
Andrew Schoen [Wed, 12 Apr 2017 12:55:03 +0000 (07:55 -0500)]
ceph-common: is ceph_rchs is True do not include install_on_redhat.yml

That task includes logic for upstream installs that we do not want to
run when deploying RHCS.

Signed-off-by: Andrew Schoen <aschoen@redhat.com>
(cherry picked from commit 37d38b122bde02efc1bde0d61ceeee3999ebb7ff)

8 years agotests: set needed config in group_vars/all for rhcs testing
Andrew Schoen [Wed, 12 Apr 2017 12:53:49 +0000 (07:53 -0500)]
tests: set needed config in group_vars/all for rhcs testing

Instead of relying on environment variables and --extra-vars simply
modify the group_vars/all that ships with the specific testing scenario
to enable ceph_rchs testing.

Signed-off-by: Andrew Schoen <aschoen@redhat.com>
(cherry picked from commit 386afbec83b72dcc3a552682d4376d1bf6a99417)

Conflicts:
tests/functional/rhcs_setup.yml

8 years agotests: remove CEPH_STABLE environment variable
Andrew Schoen [Wed, 12 Apr 2017 12:50:41 +0000 (07:50 -0500)]
tests: remove CEPH_STABLE environment variable

I'm removing this because when we use an 'rhcs' scenario then we attempt
to set CEPH_STABLE=false as an environment variable. The issue with that
is because the value is coming from an environment variable it is always
treated as a string and ansible treats that as a boolean True. I plan to
set the ceph_stable value with our rhcs_setup.yml playbook instead of
relying on ---extra-vars and environment variables.

Related ansible issue: https://github.com/ansible/ansible/issues/17193

Signed-off-by: Andrew Schoen <aschoen@redhat.com>
(cherry picked from commit 63983fa0b728571518ecdeca249ef37506a00a64)

8 years agoMerge pull request #1425 from ceph/bump-kraken v2.2.0
Sébastien Han [Mon, 10 Apr 2017 17:03:39 +0000 (19:03 +0200)]
Merge pull request #1425 from ceph/bump-kraken

common: bump ceph version to kraken

8 years agocommon: bump ceph version to kraken 1425/head
Sébastien Han [Mon, 10 Apr 2017 14:20:49 +0000 (16:20 +0200)]
common: bump ceph version to kraken

Kraken has been out for a couple of weeks now and the CI can test both
Kraken and Jewel.

Signed-off-by: Sébastien Han <seb@redhat.com>
8 years agoMerge pull request #1377 from ceph/ceph-mgr
Sébastien Han [Mon, 10 Apr 2017 15:50:24 +0000 (17:50 +0200)]
Merge pull request #1377 from ceph/ceph-mgr

mgr: add new role for ceph-mgr

8 years agotest: add mgr bootstrap support 1377/head
Sébastien Han [Thu, 6 Apr 2017 15:26:52 +0000 (17:26 +0200)]
test: add mgr bootstrap support

Signed-off-by: Sébastien Han <seb@redhat.com>
8 years agodocker: fix monitors name
Sébastien Han [Tue, 28 Mar 2017 12:15:27 +0000 (14:15 +0200)]
docker: fix monitors name

Prior to this change we were deploying a monitor using tis fqdn name but
we were checking its state and performing actions on it using its
shortname.

Signed-off-by: Sébastien Han <seb@redhat.com>
8 years agomgr: add new role for ceph-mgr
Sébastien Han [Thu, 16 Mar 2017 09:17:08 +0000 (10:17 +0100)]
mgr: add new role for ceph-mgr

The Ceph Manager daemon (ceph-mgr) runs alongside monitor daemons, to
provide additional monitoring and interfaces to external monitoring and
management systems.

Only works as of the Kraken release.

Co-Authored-By: Guillaume Abrioux <gabrioux@redhat.com>
Signed-off-by: Sébastien Han <seb@redhat.com>
8 years agoMerge pull request #1421 from ceph/tox-docs
Alfredo Deza [Mon, 10 Apr 2017 11:56:40 +0000 (07:56 -0400)]
Merge pull request #1421 from ceph/tox-docs

tox-driven docs

8 years agoMerge pull request #1422 from k0ste/newfeature
Sébastien Han [Mon, 10 Apr 2017 09:07:39 +0000 (11:07 +0200)]
Merge pull request #1422 from k0ste/newfeature

ceph-common: persist "disable_transparent_hugepage"

8 years agoceph-common: disable_transparent_hugepage now persist 1422/head
Konstantin Shalygin [Sat, 8 Apr 2017 06:06:18 +0000 (13:06 +0700)]
ceph-common: disable_transparent_hugepage now persist

8 years agodocs: add missing sections/refs in scenarios.rst 1421/head
Alfredo Deza [Fri, 7 Apr 2017 16:18:15 +0000 (12:18 -0400)]
docs: add missing sections/refs in scenarios.rst

Signed-off-by: Alfredo Deza <adeza@redhat.com>
8 years agodocs: include testing conventions and testinfra in layout.rst
Alfredo Deza [Fri, 7 Apr 2017 16:17:53 +0000 (12:17 -0400)]
docs: include testing conventions and testinfra in layout.rst

Signed-off-by: Alfredo Deza <adeza@redhat.com>
8 years agodocs: fix typo in testing/index.rst
Alfredo Deza [Fri, 7 Apr 2017 16:17:31 +0000 (12:17 -0400)]
docs: fix typo in testing/index.rst

Signed-off-by: Alfredo Deza <adeza@redhat.com>
8 years agodocs: include the tests and tox files in the glossary
Alfredo Deza [Fri, 7 Apr 2017 16:17:13 +0000 (12:17 -0400)]
docs: include the tests and tox files in the glossary

Signed-off-by: Alfredo Deza <adeza@redhat.com>
8 years agodocs: create a tests doc file
Alfredo Deza [Fri, 7 Apr 2017 16:07:42 +0000 (12:07 -0400)]
docs: create a tests doc file

Signed-off-by: Alfredo Deza <adeza@redhat.com>
8 years agodocs: create a separate tox doc file
Alfredo Deza [Fri, 7 Apr 2017 16:07:26 +0000 (12:07 -0400)]
docs: create a separate tox doc file

Signed-off-by: Alfredo Deza <adeza@redhat.com>
8 years agodocs: use glossary for the :doc: directive
Alfredo Deza [Fri, 7 Apr 2017 16:06:36 +0000 (12:06 -0400)]
docs: use glossary for the :doc: directive

Signed-off-by: Alfredo Deza <adeza@redhat.com>
8 years agodocs: add the testing glossary in the toctree
Alfredo Deza [Fri, 7 Apr 2017 16:06:16 +0000 (12:06 -0400)]
docs: add the testing glossary in the toctree

Signed-off-by: Alfredo Deza <adeza@redhat.com>
8 years agodocs: normalize the glossary vs. contents files
Alfredo Deza [Fri, 7 Apr 2017 16:05:55 +0000 (12:05 -0400)]
docs: normalize the glossary vs. contents files

Signed-off-by: Alfredo Deza <adeza@redhat.com>
8 years agodocs: create a separate docs tox.ini
Alfredo Deza [Fri, 7 Apr 2017 15:41:59 +0000 (11:41 -0400)]
docs: create a separate docs tox.ini

Signed-off-by: Alfredo Deza <adeza@redhat.com>
8 years agoMerge pull request #1412 from guits/fix_handlers
Sébastien Han [Fri, 7 Apr 2017 00:36:27 +0000 (02:36 +0200)]
Merge pull request #1412 from guits/fix_handlers

Common: Fix handlers that are not properly triggered.

8 years agoMerge pull request #1416 from ceph/WingkaiHo-master
Sébastien Han [Fri, 7 Apr 2017 00:36:17 +0000 (02:36 +0200)]
Merge pull request #1416 from ceph/WingkaiHo-master

enable ceph-mon.target service after monitor start

8 years agoMerge pull request #1418 from ceph/docs
Andrew Schoen [Thu, 6 Apr 2017 20:05:40 +0000 (15:05 -0500)]
Merge pull request #1418 from ceph/docs

Docs: an initial take on only test documentation

8 years agodocs: initial structure for tests docs 1418/head
Alfredo Deza [Thu, 6 Apr 2017 17:17:35 +0000 (13:17 -0400)]
docs: initial structure for tests docs

Signed-off-by: Alfredo Deza <adeza@redhat.com>
8 years agodocs create a separate glossary file
Alfredo Deza [Thu, 6 Apr 2017 17:14:45 +0000 (13:14 -0400)]
docs create a separate glossary file

Signed-off-by: Alfredo Deza <adeza@redhat.com>
8 years agodocs: the master doc is the glossary file
Alfredo Deza [Thu, 6 Apr 2017 17:10:54 +0000 (13:10 -0400)]
docs: the master doc is the glossary file

Signed-off-by: Alfredo Deza <adeza@redhat.com>
8 years agodocs: ignore build contents from git
Alfredo Deza [Wed, 5 Apr 2017 16:27:25 +0000 (12:27 -0400)]
docs: ignore build contents from git

Signed-off-by: Alfredo Deza <adeza@redhat.com>
8 years agodocs: initial take on doc structure
Alfredo Deza [Wed, 5 Apr 2017 16:27:06 +0000 (12:27 -0400)]
docs: initial take on doc structure

Signed-off-by: Alfredo Deza <adeza@redhat.com>
8 years agoMerge pull request #1417 from ceph/docker-env-vars
Andrew Schoen [Thu, 6 Apr 2017 17:11:16 +0000 (12:11 -0500)]
Merge pull request #1417 from ceph/docker-env-vars

tests: add docker environment variables to tox.ini

8 years agotests: add docker environment variables to tox.ini 1417/head
Andrew Schoen [Thu, 6 Apr 2017 15:29:16 +0000 (10:29 -0500)]
tests: add docker environment variables to tox.ini

This allows us to control the docker registry, image name and tag name
for the containers.

Signed-off-by: Andrew Schoen <aschoen@redhat.com>
8 years agoenable ceph-mon.target service after monitor start 1416/head
WingkaiHo [Wed, 29 Mar 2017 03:11:04 +0000 (11:11 +0800)]
enable ceph-mon.target service after monitor start

ceph-create-keys unit file was removed here:

* https://github.com/ceph/ceph/commit/8bcb4646b6b9846bb965cdec3ca2a21eb3b26bab
* https://github.com/ceph/ceph/commit/dc5fe8d415858358bd0baf5d8dce0a753f5e0cea

As a consequence the systemctl preset command now fails to run since the
unit does not exist anymore. Due to the redirection in /dev/null we
don't know what's happening.

Ultimately the mon unit doesn't get enabled and the mon service won't
start after reboot.
Removing the old/non-existent unit makes the command succeed now.

ceph fix: https://github.com/ceph/ceph/pull/14226

Signed-off-by: WingkaiHo <sanguosfiang@163.com>
Co-Authored-By: Sébastien Han <seb@redhat.com>
8 years agoCommon: Fix handlers that are not properly triggered. 1412/head
Guillaume Abrioux [Mon, 3 Apr 2017 17:55:11 +0000 (19:55 +0200)]
Common: Fix handlers that are not properly triggered.

Until now, only the first task were executed.
The idea here is to use `listen` statement to be able to notify multiple
handler and regroup all of them in `./handlers/main.yml` as notifying an
included handler task is not possible.

Signed-off-by: Guillaume Abrioux <gabrioux@redhat.com>
8 years agoMerge pull request #1415 from dahorak/osd-autodiscovery-fix-check-holders
Sébastien Han [Thu, 6 Apr 2017 08:56:42 +0000 (10:56 +0200)]
Merge pull request #1415 from dahorak/osd-autodiscovery-fix-check-holders

osd autodiscovery mode: fix holders detection

8 years agoosd autodiscovery mode: fix holders detection 1415/head
Daniel Horak [Thu, 6 Apr 2017 07:11:32 +0000 (09:11 +0200)]
osd autodiscovery mode: fix holders detection

Small fix for (probably copy&paste) issue from 42ffe6301.

Signed-off-by: Daniel Horak <dahorak@redhat.com>
8 years agoMerge pull request #1414 from ceph/scenario-alias
Alfredo Deza [Wed, 5 Apr 2017 21:42:06 +0000 (17:42 -0400)]
Merge pull request #1414 from ceph/scenario-alias

tests: add an alias of cluster for the centos7_cluster scenario

8 years agotests: add an alias of cluster for the centos7_cluster scenario 1414/head
Andrew Schoen [Wed, 5 Apr 2017 18:36:51 +0000 (13:36 -0500)]
tests: add an alias of cluster for the centos7_cluster scenario

When testing this downstream it makes more sense for this scenario to be
named just 'cluster' because have 'centos7' in the name is misleading.

Signed-off-by: Andrew Schoen <aschoen@redhat.com>
8 years agoMerge pull request #1411 from ceph/bump-ansible
Sébastien Han [Wed, 5 Apr 2017 14:03:20 +0000 (16:03 +0200)]
Merge pull request #1411 from ceph/bump-ansible

tests: bump the version of ansible to 2.2.2

8 years agoMerge pull request #1409 from ceph/auto-dm
Sébastien Han [Wed, 5 Apr 2017 13:58:41 +0000 (15:58 +0200)]
Merge pull request #1409 from ceph/auto-dm

osd: autodiscovery mode, use holders to detect device

8 years agoMerge pull request #1410 from ceph/kraken-tests
Sébastien Han [Wed, 5 Apr 2017 11:53:17 +0000 (13:53 +0200)]
Merge pull request #1410 from ceph/kraken-tests

tests: run all tests for jewel and kraken

8 years agotests: bump the version of ansible to 2.2.2 1411/head
Alfredo Deza [Wed, 5 Apr 2017 11:38:58 +0000 (07:38 -0400)]
tests: bump the version of ansible to 2.2.2

Signed-off-by: Alfredo Deza <adeza@redhat.com>
8 years agoosd: autodiscovery mode, use holders to detect device 1409/head
Sébastien Han [Tue, 4 Apr 2017 08:33:22 +0000 (10:33 +0200)]
osd: autodiscovery mode, use holders to detect device

As reported in
https://github.com/ceph/ceph-ansible/issues/1403 when devices are held
by lvm and `osd_auto_discovery` is set to true, it's not enough to check
for a partition count = 0 since Ansible does not report.
This patch also looks for 'holders' which in a case of lvm corresponds
to the name of the pv. Now we also look for holders = 0.

Fixes: #1403
Signed-off-by: Sébastien Han <seb@redhat.com>
8 years agotests: run all tests for jewel and kraken 1410/head
Andrew Schoen [Mon, 3 Apr 2017 21:11:55 +0000 (16:11 -0500)]
tests: run all tests for jewel and kraken

Signed-off-by: Andrew Schoen <aschoen@redhat.com>
8 years agoMerge pull request #1404 from ceph/check-osd-id
Sébastien Han [Thu, 30 Mar 2017 12:52:58 +0000 (14:52 +0200)]
Merge pull request #1404 from ceph/check-osd-id

playbook: homogenize the way list osd ids

8 years agoplaybook: homogenize the way list osd ids 1404/head
Sébastien Han [Thu, 30 Mar 2017 09:51:38 +0000 (11:51 +0200)]
playbook: homogenize the way list osd ids

Problem: too many different commands to do the same thing. The 'cut'
command on infrastructure-playbooks/purge-cluster.yml was also wrong.
This sed command from osixia in ceph-docker
https://github.com/ceph/ceph-docker/pull/580/ addresses all the
scenarios.

Signed-off-by: Sébastien Han <seb@redhat.com>
8 years agoMerge pull request #1399 from guits/fix_ntp_install
Sébastien Han [Wed, 29 Mar 2017 09:50:27 +0000 (11:50 +0200)]
Merge pull request #1399 from guits/fix_ntp_install

Common: Do not install ntp when ntp_service_enabled is false

8 years agoCommon: Do not install ntp when ntp_service_enabled is false 1399/head
Guillaume Abrioux [Tue, 28 Mar 2017 15:45:09 +0000 (17:45 +0200)]
Common: Do not install ntp when ntp_service_enabled is false

ntp is still installed even if ntp_service_enabled is set to false.
That could be a problem if the time synchronization is managed by
something else than ceph-ansible or if you want to use different NTP
implementation as suggested in #1354.

Fixes: #1354
Signed-off-by: Guillaume Abrioux <gabrioux@redhat.com>
Signed-off-by: Guits <gabrioux@redhat.com>
8 years agoMerge pull request #1397 from ceph/debug
Sébastien Han [Tue, 28 Mar 2017 10:15:49 +0000 (12:15 +0200)]
Merge pull request #1397 from ceph/debug

vagrant: add debug option

8 years agovagrant: add debug option 1397/head
Sébastien Han [Tue, 28 Mar 2017 10:00:32 +0000 (12:00 +0200)]
vagrant: add debug option

If debug is set to true in vagrant_variables.yml then during the vagrant
provision phase Ansible will run with -vvvv option

Signed-off-by: Sébastien Han <seb@redhat.com>
8 years agoMerge pull request #1386 from WingkaiHo/master
Sébastien Han [Tue, 28 Mar 2017 07:50:39 +0000 (09:50 +0200)]
Merge pull request #1386 from WingkaiHo/master

Create recover-osds-after-ssd-journal-failure.yml

8 years agoMerge pull request #1391 from k0ste/newfeature
Sébastien Han [Mon, 27 Mar 2017 16:48:33 +0000 (18:48 +0200)]
Merge pull request #1391 from k0ste/newfeature

Resolve issues when groups names not in default value.

8 years agoMerge pull request #1396 from guits/fix_1374
Sébastien Han [Mon, 27 Mar 2017 16:13:51 +0000 (18:13 +0200)]
Merge pull request #1396 from guits/fix_1374

BugFix: Fix #1374

8 years agoResolve issues when groups names not in default value. 1391/head
Konstantin Shalygin [Fri, 24 Mar 2017 06:40:15 +0000 (13:40 +0700)]
Resolve issues when groups names not in default value.

8 years agoBugFix: Fix #1374 1396/head
Guillaume Abrioux [Thu, 23 Mar 2017 15:00:02 +0000 (16:00 +0100)]
BugFix: Fix #1374

If a group of hosts is empty, (for instance 'mdss', in case of a
deployment without any mds node), the playbook will fails when trying
to restart service with `"'dict object' has no attribute u'XXX'"` error.

The idea here is to force the `with_items` statements in all included handler tasks
to get at least an empty array.

Signed-off-by: Guillaume Abrioux <gabrioux@redhat.com>
8 years agoMerge pull request #1394 from concubidated/master
Sébastien Han [Mon, 27 Mar 2017 09:12:15 +0000 (11:12 +0200)]
Merge pull request #1394 from concubidated/master

ceph-common: update sysctl file location

8 years agoremove the change 1386/head
WingkaiHo [Mon, 27 Mar 2017 03:06:32 +0000 (11:06 +0800)]
remove the change

8 years agoenable ceph-mon.target service after monitor start, fix the ceph monitor service...
WingkaiHo [Mon, 27 Mar 2017 03:01:35 +0000 (11:01 +0800)]
enable ceph-mon.target service after monitor start, fix the ceph monitor service do not start when the ceph reboot

8 years agoMerge https://github.com/ceph/ceph-ansible
WingkaiHo [Mon, 27 Mar 2017 02:50:38 +0000 (10:50 +0800)]
Merge https://github.com/ceph/ceph-ansible

8 years agoceph-common: update sysctl file location 1394/head
Concubidated [Fri, 24 Mar 2017 19:52:37 +0000 (12:52 -0700)]
ceph-common: update sysctl file location

systctl tuning should be in the sysctl.d directory. This creates
a seperation from what values were set specific to ceph, and what
values were set by the operator.

Signed-off-by: Tyler Brekke <tbrekke@redhat.com>
8 years agoMerge pull request #1389 from ceph/downstream-testing
Andrew Schoen [Fri, 24 Mar 2017 15:09:12 +0000 (10:09 -0500)]
Merge pull request #1389 from ceph/downstream-testing

tests: enable testing of rhcs

8 years agotests: change ceph/ubuntu-xenial boxes to rhel7 1389/head
Andrew Schoen [Thu, 23 Mar 2017 13:43:11 +0000 (08:43 -0500)]
tests: change ceph/ubuntu-xenial boxes to rhel7

Signed-off-by: Andrew Schoen <aschoen@redhat.com>
8 years agotests: when testing rhcs default CEPH_STABLE=false
Andrew Schoen [Thu, 23 Mar 2017 11:42:43 +0000 (06:42 -0500)]
tests: when testing rhcs default CEPH_STABLE=false

This will keep ceph-ansible from installing the upstream repo file on
the testing nodes.

Signed-off-by: Andrew Schoen <aschoen@redhat.com>
8 years agocommand module instead shell
WingkaiHo [Thu, 23 Mar 2017 09:38:41 +0000 (17:38 +0800)]
command module instead shell

8 years agoremove service enable
WingkaiHo [Thu, 23 Mar 2017 07:28:14 +0000 (15:28 +0800)]
remove service enable

8 years agobreak it into lines since you already use the string block synta and fix disable...
WingkaiHo [Thu, 23 Mar 2017 06:49:10 +0000 (14:49 +0800)]
break it into lines since you already use the string block synta and fix disable it here and enable again in later task

8 years agoremove this detection and simply rely on {{ cluster }}
WingkaiHo [Thu, 23 Mar 2017 01:22:06 +0000 (09:22 +0800)]
remove this detection and simply rely on {{ cluster }}

8 years agotests: change hosts in first play in rhcs_setup to localhost
Andrew Schoen [Thu, 23 Mar 2017 00:40:34 +0000 (19:40 -0500)]
tests: change hosts in first play in rhcs_setup to localhost

Signed-off-by: Andrew Schoen <aschoen@redhat.com>
8 years agotests: set MTU to 1400 on test node interfaces
Andrew Schoen [Wed, 22 Mar 2017 21:44:29 +0000 (16:44 -0500)]
tests: set MTU to 1400 on test node interfaces

In the environment we were testing on, MTU was set to 1500 which causes
download failures of our yum repos. There might be a better way to set
this instead of doing it here in ansible.

Signed-off-by: Andrew Schoen <aschoen@redhat.com>