Commit Graph

3616 Commits (08a2b58d39a687e25436afdf3fda1591d3be8ca1)
 

Author SHA1 Message Date
Sébastien Han 3ce0621b44 ceph-osd: use fact logic to set permissions
Signed-off-by: Sébastien Han <seb@redhat.com>
2016-02-21 18:27:36 +01:00
Sébastien Han 17fb38294e Octal file permissions must contain leading zero
Signed-off-by: Sébastien Han <seb@redhat.com>
2016-02-21 18:21:19 +01:00
Leseb dba88502b0 Merge pull request #551 from ceph/add-ceph-fetch-key
contrib: add ceph-fetch-keys
2016-02-18 18:08:27 +01:00
Sébastien Han 077d9e0f71 contrib: add ceph-fetch-keys
Signed-off-by: Sébastien Han <seb@redhat.com>
2016-02-18 18:07:43 +01:00
Leseb 82e9aac9ea Merge pull request #550 from ceph/fetch-keys-role
Adds a new role, ceph-fetch-keys
2016-02-18 17:52:57 +01:00
Andrew Schoen a69091505e a playbook to add an OSD to an existing cluster
The playbook uses the ceph-fetch-keys role to connect to the mons and
retrieve keys for the OSD bootstrapping. Ensure that all your mons for
the cluster you're adding the OSD to is in your inventory in the mons
group. This ensures that a proper ceph.conf is created for the new OSD.

Signed-off-by: Andrew Schoen <aschoen@redhat.com>
2016-02-18 09:52:26 -06:00
Andrew Schoen 6fa25fa057 Adds a new ceph-fetch-keys role
The purpose of this is so we can connect to the mons and gather the keys
needed to configure an OSD or additonal MON without having to reconfigure
the existing mons at the same time.

Signed-off-by: Andrew Schoen <aschoen@redhat.com>
2016-02-18 09:52:26 -06:00
Leseb e3dd88a725 Merge pull request #549 from retspen/master
Added support parallels provider
2016-02-18 13:58:30 +01:00
retspen 988010f2b8 Added support parallels provider 2016-02-18 13:58:33 +02:00
Leseb 579d09c065 Merge pull request #548 from ceph/more-tags
adds some more --package-install tags; make check_mandatory_vars more specific
2016-02-17 10:26:18 +01:00
Andrew Schoen bcf34d539f make mandatory variable checks only apply when needed
In our use case we might only be configuring mons and not osds in the
same call, so we don't want to check variables needed for osds when they
are not needed to configure a mon.

Signed-off-by: Andrew Schoen <aschoen@redhat.com>
2016-02-16 10:43:36 -06:00
Andrew Schoen beb966421e add the package-install tags to some tasks in check_mandatory_vars
We'll want these checks to always run when we're installing packages

Signed-off-by: Andrew Schoen <aschoen@redhat.com>
2016-02-16 10:43:36 -06:00
Leseb f86da22ce6 Merge pull request #543 from ceph/min_free_kbytes
ceph-common: calculate vm.min_free_kbytes
2016-02-16 16:27:23 +01:00
Sébastien Han 68a9d45a49 ceph-common: use stdout for the register var
Signed-off-by: Sébastien Han <seb@redhat.com>
2016-02-16 16:26:56 +01:00
Sébastien Han 7ea9d95815 ceph-common: do not calculate vm.min_free_kbytes
we preserve the default kernel value.

Signed-off-by: Sébastien Han <seb@redhat.com>
2016-02-16 15:40:49 +01:00
Leseb e2e99bfb79 Merge pull request #542 from KGoreczny/upgrade_purge_cluster
Rewrite purge-cluster playbook
2016-02-16 14:10:10 +01:00
Leseb 349814d101 Merge pull request #546 from ceph/separate-agent
separate the agent so it doesn't break standard installs
2016-02-16 14:08:26 +01:00
KGoreczny 6e03f4944b Unify quotation marks
Signed-off-by: KGoreczny <krzysztof.goreczny@intel.com>
2016-02-16 13:41:58 +01:00
Alfredo Deza a28944c46a remove the agent from the mons
Signed-off-by: Alfredo Deza <adeza@redhat.com>
2016-02-16 07:12:07 -05:00
Alfredo Deza a8358a0349 separate the agent so it doesn't break standard installs
Mainly because some packages are still not there for upstream which will cause
it to break installs that would otherwise would succeed

Signed-off-by: Alfredo Deza <adeza@redhat.com>
2016-02-16 07:09:54 -05:00
Leseb 9db28f52ed Merge pull request #544 from ceph/contrib-rhscon
contrib: add the ceph-agent
2016-02-15 17:43:37 +01:00
Leseb 7c33a27b06 Merge pull request #539 from ceph/rhscon-support
rhscon (storage agent)
2016-02-15 17:43:25 +01:00
Alfredo Deza 6ae743702f gitignore group_vars/agent
Signed-off-by: Alfredo Deza <adeza@redhat.com>
2016-02-15 11:42:01 -05:00
Sébastien Han 3c074fba9b contrib: add the ceph-agent
Signed-off-by: Sébastien Han <seb@redhat.com>
2016-02-15 17:11:03 +01:00
KGoreczny 4e2303fce5 Add dnf package support; change condition checks for package managers
Signed-off-by: KGoreczny <krzysztof.goreczny@intel.com>
2016-02-15 16:51:16 +01:00
KGoreczny fc2ace7ba0 Use yum and apt modules instead of package to provide Ansible <2.0 support; fix typo
Signed-off-by: KGoreczny <krzysztof.goreczny@intel.com>
2016-02-15 16:01:00 +01:00
Sébastien Han 6a3c4311b5 ceph-common: update group_vars vm.min_free_kbytes
Signed-off-by: Sébastien Han <seb@redhat.com>
2016-02-15 15:05:42 +01:00
Leseb 270d98bb16 Merge pull request #541 from haasn/master
ceph-common: use explicit action syntax for config_template
2016-02-15 15:04:37 +01:00
Sébastien Han 4ede7211a0 ceph-common: calculate vm.min_free_kbytes
based on the os memory we calculate the value for vm.min_free_kbytes

Signed-off-by: Sébastien Han <seb@redhat.com>
2016-02-15 15:03:10 +01:00
Krzysztof Goreczny bbcdf2fde7 Rewrite purge-cluster playbook
Signed-off-by: Krzysztof Goreczny <krzysztof.goreczny@intel.com>
2016-02-15 14:22:25 +01:00
Alfredo Deza b0fce503a8 use the agent on monitors on site.yml.sample
Signed-off-by: Alfredo Deza <adeza@redhat.com>
2016-02-15 07:58:02 -05:00
Alfredo Deza f49b348979 create a group_vars agent sample file
Signed-off-by: Alfredo Deza <adeza@redhat.com>
2016-02-15 07:52:46 -05:00
Alfredo Deza da4b7b290c fix wording on defaults main.yml
Signed-off-by: Alfredo Deza <adeza@redhat.com>
2016-02-15 07:50:09 -05:00
Alfredo Deza 73928a5d7b only check for subscription manager on RH systems
Signed-off-by: Alfredo Deza <adeza@redhat.com>
2016-02-15 07:49:03 -05:00
Niklas Haas 20e6f932dd
ceph-common: use explicit action syntax for config_template
This fails with "ERROR! no action detected in task" otherwise.
2016-02-14 20:58:04 +01:00
Leseb 3ba68d3836 Merge pull request #540 from ceph/fix-autodiscover-osd
ceph-osd: fix the autodiscovery osd scenario
2016-02-13 23:19:50 +01:00
Sébastien Han 59b96e9f11 ceph-osd: fix the autodiscovery osd scenario
the parted command wasn't getting the devices properly and the partition
count is not necessary.

Signed-off-by: Sébastien Han <seb@redhat.com>
2016-02-13 23:07:44 +01:00
Alfredo Deza c59d6f00cb check for subscription manager
Signed-off-by: Alfredo Deza <adeza@redhat.com>
2016-02-12 15:54:56 -05:00
Alfredo Deza 2fd5ab18ac create the task to install and configure the agent
Signed-off-by: Alfredo Deza <adeza@redhat.com>
2016-02-12 15:37:06 -05:00
Alfredo Deza 2d0993d4c7 better reword the master host variable
Signed-off-by: Alfredo Deza <adeza@redhat.com>
2016-02-12 13:27:04 -05:00
Alfredo Deza 42e5391723 create a template for the ceph agent minion
Signed-off-by: Alfredo Deza <adeza@redhat.com>
2016-02-12 13:08:32 -05:00
Alfredo Deza 28abc3d6d4 define the supported versions
Signed-off-by: Alfredo Deza <adeza@redhat.com>
2016-02-12 12:01:51 -05:00
Alfredo Deza b4dc6cf43f initial files for the ceph-agent role
Signed-off-by: Alfredo Deza <adeza@redhat.com>
2016-02-12 11:53:15 -05:00
Leseb 46f84c03fa Merge pull request #538 from ceph/galaxy-requirements
add a galaxy dependency file example
2016-02-12 15:56:30 +01:00
Sébastien Han b6d734e43b add a galaxy dependency file example
Signed-off-by: Sébastien Han <seb@redhat.com>
2016-02-12 15:55:42 +01:00
Leseb eaee28093e Merge pull request #536 from ceph/fix-auto-discover-disk
ceph-osd: fix the auto discovery scenario
2016-02-12 01:08:09 +01:00
Sébastien Han 64c458bfcf ceph-osd: fix register variable
as stated in https://github.com/ansible/ansible/issues/4297
if we register a variable twice and even if a task is skipped the
  register will not get overwritten... So we use the fact variant as
  mentionned in the ansible issue.

Signed-off-by: Sébastien Han <seb@redhat.com>
2016-02-12 00:54:09 +01:00
Sébastien Han b285a5ae8f ceph-osd: simply the regex to check the device
Signed-off-by: Sébastien Han <seb@redhat.com>
2016-02-11 22:51:23 +01:00
Sébastien Han ea0979cbbe ceph-osd: fix the auto discovery scenario
While this is not widly used (AFAIK :p) the feature was broken. Thanks
to @zmc for reporting it. You can now set `osd_auto_discovery` to
true in your group_vars/osd and it will go through all the devices
available and will make them OSDs.

Signed-off-by: Sébastien Han <seb@redhat.com>
2016-02-11 22:43:09 +01:00
Leseb 685b5f2940 Merge pull request #535 from dang/rgw_addr
Allow RGW and client VMs together
2016-02-11 21:25:14 +01:00