x
/
fuel-ccp-installer
Code Issues Proposed changes
552 Commits 1 Branch 2 Tags 2.9 MiB
Commit Graph

150 Commits

Author SHA1 Message Date
Andreas Jaeger 5e717e6568 Retire repository 
Fuel (from openstack namespace) and fuel-ccp (in x namespace)
repositories are unused and ready to retire.

This change removes all content from the repository and adds the usual
README file to point out that the repository is retired following the
process from
https://docs.openstack.org/infra/manual/drivers.html#retiring-a-project

See also
http://lists.openstack.org/pipermail/openstack-discuss/2019-December/011647.html

Depends-On: https://review.opendev.org/699362
Change-Id: Icaff65f27554a379e9130bf9354b27ce0eb3b2a6
 2019-12-18 09:47:51 +01:00
Jenkins e751b92d75 Merge "Update ansible to 2.3.0" 2017-07-28 17:04:07 +00:00
Matthew Mosesohn 51164d9aab switch back to python2 
Change-Id: I606bfd79db64370fc60352cad113e81259cbd254
 2017-07-28 19:47:28 +03:00
Matthew Mosesohn 0652e1b6d3 install python3-netaddr 
Change-Id: I3cfc56ae9f7cfe3c74705910b1f2ed27c51ef485
 2017-07-28 15:15:08 +03:00
Matthew Mosesohn 169f2ed141 use python-pip3 
Change-Id: Ide9204a68f8f4e1bb67c8bd2a2f8c9c3e1fb9f45
 2017-07-27 20:18:32 +03:00
Matthew Mosesohn e0f6a8bd4e workaround for pyasn1 issues with pip 
Change-Id: Ia89eb5894bb6ce28c3b25d0ee949d5346d656813
 2017-07-27 19:47:07 +03:00
Sergii Golovatiuk eeccc79109 Update ansible to 2.3.0 
Change-Id: If05c546be0922fc9e9e4dfccefc751a1e4285cf1
 2017-04-27 12:52:12 +02:00
Matthew Mosesohn 05f11e6348 Improve tolerance for failed nodes 
Raise SSH reconnects to 20
Disable any_errors_fatal

Depends on https://github.com/kubernetes-incubator/kargo/pull/1201/

Change-Id: I0d20c9e0e80f87561c50957fdff1e576ec89646c
 2017-04-04 13:46:18 +03:00
Matthew Mosesohn 06591089f6 Add scale mode defaults 
Enable scale by deploying over 50 nodes or explicitly set SCALE_MODE=yes

Change-Id: Icb19585b0cb1fa76550c3188ca5a0b27b37a759a
 2017-03-01 16:26:50 +03:00
Matthew Mosesohn 759a0f3abe Symlink group_vars directory instead of use as explicit varfile 
This allows proper group var default value inheritance, instead
of applying group vars to all roles. Secondly, it reduces the
amount of management as Kargo introduces more var files for groups.

Also fixed bin_dir for configure_lgos role.

Change-Id: Iadb44d67c90a0af9bf5403395be6d91fbe9e9ef1
 2017-02-28 13:45:26 +03:00
Sergii Golovatiuk 0667f39786 Fix fuel-ccp-installer 
- Include k8s-cluster.yml that was introduced by recent kargo changes
- Fix typo
- Move user below to have more clear ansible-playbook line

Change-Id: I008b7d124ef651263980b191f07b1d4daec34972
 2017-02-24 14:12:40 +01:00
Matthew Mosesohn 120066b34c Pin Ansible to 2.2.1 
Recent changes were made to Kargo cluster.yml which require
Ansible 2.2.1. Kargo is fully compatible with this version of
Ansible already.

Change-Id: I3db7d210cf5038809d4674a709f9ee63633b0a48
 2017-02-09 16:23:17 +03:00
Jenkins 8c9962e10c Merge "Remove inventory repo support" 2017-02-07 17:09:55 +00:00
Jenkins 7026bb045f Merge "Install python-pip, ensure other admin node packages" 2017-02-02 14:50:50 +00:00
Matthew Mosesohn 4558daaf51 Install python-pip, ensure other admin node packages 
Now ensures sshpass and git on admin node

Change-Id: I28802cda3b66c2d8a0949b63575f89591160edc5
 2017-02-02 17:22:01 +03:00
Matthew Mosesohn 7be422c171 Remove inventory repo support 
External inventory repo support adds a lot of baggage to project scope.
The new workflow is a bit simpler:
External commands:
git clone fuel-ccp-installer
git clone inventory_repo inventory
run kargo_deploy.sh

Behavior is now reverted in that if inventory/ dir exists, it will not
overwrite default file contents. However, if CUSTOM_YAML is specified,
it will overwrite data.

Change-Id: I9557cf86972a64ba0e72aa86111f598be66caff3
 2017-02-01 18:26:32 +03:00
Matthew Mosesohn f698d3396a Set bridge-nf-call-iptables to 0 for fuel-devops deploy 
Change-Id: I5cb855a3864d7fe5dd538de40155fd5da2d89aa2
 2017-01-31 20:16:57 +03:00
Sergii Golovatiuk 86e13dc310 Reduce number of ssh connections 
Since we implemented multiline admin_node_command the number of
unnecessary ssh connections may be reduced.

Change-Id: Ia8f37b565ecc8448dca847a47eb68803ed535fe7
 2017-01-20 16:58:18 +01:00
Jenkins 0ea999d996 Merge "Fix admin_node_command stdin handling" 2017-01-20 14:06:51 +00:00
Matthew Mosesohn de2459aa99 Fix admin_node_command stdin handling 
Change-Id: Iac3fea907f59539adced2ffd4db565edfd7da47e
 2017-01-19 13:22:37 +03:00
Matthew Mosesohn 6fdac9e4cc Install ansible from pip by default 
Legacy behavior can be enabled by running
export ANSIBLE_INSTALL_SOURCE=apt

Change-Id: I9fed27340ec8869c6459792ffaff13f77473c738
 2017-01-19 09:30:55 +03:00
Artur Zarzycki 2d517fccf1 Revert "Change ansible version to 2.2.0" 
This reverts commit 44e4e723cd.

Change-Id: Ib66221fc3d97a8618263d38ba303ddaf747cb992
 2017-01-18 08:14:09 +00:00
Artur Zarzycki 44e4e723cd Change ansible version to 2.2.0 
We don't want to be blocked until
https://github.com/kubernetes-incubator/kargo/pull/910
will be merged.

Change-Id: I64c08a4442a6f684ce50fb590c07108cb539ad73
 2017-01-17 15:26:52 +01:00
Matthew Mosesohn 652276b02a Adapt admin_node_command to accept commands both via stdin and args 
Change-Id: I04800a9c5e24d31942864dd5554043142584f6eb
 2017-01-16 12:14:21 +01:00
Sergii Golovatiuk 47f83f7e5b Pipefail to wait_for_apt_lock_release and wait_for_apt_lock_release 
- arguments passed to with_retries may contain pipes. In this case we
  should catch issues from any command rather than the last one
- Add wait_for_apt_lock_release before every apt|dpkg invokation. This
  assures that other processes don't lock apt|dpkg files.
- moving $@ in with_ansible to the last for readability reasons
- Change (( tries-- )) to $(( tries-- )) to fix bashate E043 error
- Replace [[ $arg > 1 ]] to -gt
- Fix intendation to comply with bashate requirements

Change-Id: Icff4bb5732d755334ae2b447cb41cdb6053dd1a3
Signed-off-by: Sergii Golovatiuk <sgolovatiuk@mirantis.com>
 2017-01-12 11:55:51 +01:00
Jenkins 908aa6dca7 Merge "Avoid catching pgrep apt from ssh source to self" 2017-01-11 14:03:59 +00:00
Matthew Mosesohn f28413491d Avoid catching pgrep apt from ssh source to self 
Change-Id: I94957d362d520f9bc9bcd90c22d05c88a6f05435
Closes-Bug: #1655414
 2017-01-11 15:53:30 +03:00
Matthew Mosesohn 4a0a5138fb Add support for ironic nodelist as inventory source 
Change-Id: I5d2dcd58409429fd8a71c9f5713a67fc032bdcd2
 2017-01-10 18:53:33 +03:00
Matthew Mosesohn f808bf6faa Remove inventory builder 
This logic is now part of Kargo in contrib/ dir.

Change-Id: Id922a611500ebebd4984b06aa2e4c85a207a9941
 2017-01-10 18:10:56 +03:00
Bogdan Dobrelya 1198d82f59 Consume early DNS configuration from upstream 
Change-Id: I9a771277a587d84f17f1bf37e7b5244dd3543855
Signed-off-by: Bogdan Dobrelya <bdobrelia@mirantis.com>
 2016-12-12 11:00:33 +01:00
Jenkins 8898246295 Merge "Switch Kargo to the incubated repo" 2016-11-07 11:58:55 +00:00
Bogdan Dobrelya 273fe37112 Switch Kargo to the incubated repo 
Change-Id: I570faa1f458923ddb09effadce5795a56ed72291
Signed-off-by: Bogdan Dobrelya <bdobrelia@mirantis.com>
 2016-10-28 15:20:22 +02:00
Matthew Mosesohn 1620a79e70 Combine preinstall and dnsmasq steps 
Depends on https://github.com/kubespray/kargo/pull/568

Change-Id: Iab24013f2331e639b9e84b79895f7d595e227035
 2016-10-28 13:49:40 +03:00
Matthew Mosesohn fa23319995 Split master/etcd from non-master deployment and raise forks 
Performs master + etcd deployment first and then non-master nodes all
at once.

Change-Id: I0840f4685c09d95c80aa56578d716f64695dee23
 2016-10-27 08:28:26 +00:00
Matthew Mosesohn 0168b4ee2b Revert "Split master/etcd from non-master deployment and raise forks" 
This reverts commit d23c5d9b54.

Change-Id: I784924f789c4f060826cdd6e8b540cf135be89aa
 2016-10-27 06:44:52 +00:00
Matthew Mosesohn d23c5d9b54 Split master/etcd from non-master deployment and raise forks 
Performs master + etcd deployment first and then non-master
nodes all at once. Raises max forks to 100 if the number of
nodes to deploy is over 50.

Change-Id: I873cb8a0b83b79bde8f9342b3e9c841427c537d9
 2016-10-26 12:37:54 +00:00
Jenkins c50d9bfd81 Merge "Enable fact caching and retries" 2016-10-24 11:11:01 +00:00
Matthew Mosesohn f8a46ec43d Enable fact caching and retries 
Add limit for retry if a retry file exists.

Change-Id: I463dd32f288c2336fc2e0aa44ecd9cd67f02a9b9
 2016-10-24 08:41:54 +00:00
Oleg Basov 971465dcd3 Setting workspace cleaning up as an option in order not to mess up with what we have from previous unsuccesful deployment. 
Change-Id: I7321f8791467fe7a5d94dc5f3f70ae90448e451c
 2016-10-20 14:49:02 +03:00
Jenkins f86b67447e Merge "Make logs collecter to honor forks/password based opts" 2016-09-19 16:47:07 +00:00
Jenkins cf8a48ced9 Merge "W/a delegate_to issues" 2016-09-19 13:31:30 +00:00
Bogdan Dobrelya 18018ffa13 Make logs collecter to honor forks/password based opts 
Change-Id: I16f325cf3efa2e44e78a070e400b408c0308355c
Signed-off-by: Bogdan Dobrelya <bdobrelia@mirantis.com>
 2016-09-19 15:05:34 +02:00
Jenkins 7d81f033ed Merge "Fixing wrong ansible forks number calculation." 2016-09-19 12:06:03 +00:00
Oleg Basov 4fefe6098c Fixing wrong ansible forks number calculation. 
Change-Id: I9b2aa923c7f68d736988e34ea3d35593658e4eb2
 2016-09-19 14:24:42 +03:00
Bogdan Dobrelya ccdbdc8701 W/a delegate_to issues 
Specify ansible_ssh_user see
https://github.com/ansible/ansible/issues/14720

Change-Id: Ic9457ea61347ca7569df3ae405c3f9bb43358316
Signed-off-by: Bogdan Dobrelya <bdobrelia@mirantis.com>
 2016-09-19 11:37:06 +02:00
Matthew Mosesohn 010a894467 Fix test for CUSTOM_YAML var being set 
Change-Id: I12b356fe43dfe844622d9f1bfe028ba8a792016d
 2016-09-19 12:28:50 +03:00
Matthew Mosesohn 427b36d90b enable custom yaml from inventory 
Change-Id: I9282b236ee7d6d66aa857b542d62557dc21fa772
 2016-09-16 20:04:10 +00:00
Matthew Mosesohn 8b51d7a4f2 Fix git review syntax error 
After deployment if there are changes made to default config,
kargo_deploy.sh will propose a gerrit change request if the
inventory repo is a gerrit repository. This is just a syntax fix.

Change-Id: I04a80fec93e512c420056aab6b52924cf4021541
 2016-09-16 14:46:24 +03:00
Jenkins 9f6826345c Merge "Raise default ansible timeout" 2016-09-14 13:56:13 +00:00
Matthew Mosesohn 9756414fde Raise default ansible timeout 
Raises timeout from 10s to 600s for scale scenarios

Change-Id: I5d3b645ec6aac311939d7798858c56cf75394796
 2016-09-14 16:16:34 +03:00