Move all pcs commands to the host

[Michele Baldessari]

Currently we create pcmk resources via especially crafted containers by running pcs commands inside. This has always been a bit constraining because pacemaker was just never built with the possibility to run these commands inside a container in mind. Every small discrepancy between host and containers keeps hurtings us (libqb CVEs, minor pacemaker changes, etc.).

We should strive to execute these commands from the host and not have to deal with these issues any longer.

[OpenStack Infra (hudson-openstack)]

Related fix proposed to branch: master
Review: https://review.opendev.org/707953

[OpenStack Infra (hudson-openstack)]

Related fix proposed to branch: master
Review: https://review.opendev.org/707954

[OpenStack Infra (hudson-openstack)]

Related fix proposed to branch: master
Review: https://review.opendev.org/707955

[OpenStack Infra (hudson-openstack)]

Related fix proposed to branch: master
Review: https://review.opendev.org/707956

[OpenStack Infra (hudson-openstack)]

Related fix proposed to branch: master
Review: https://review.opendev.org/707957

[OpenStack Infra (hudson-openstack)]

Related fix proposed to branch: master
Review: https://review.opendev.org/708357

[OpenStack Infra (hudson-openstack)]

Related fix proposed to branch: master
Review: https://review.opendev.org/708738

[OpenStack Infra (hudson-openstack)]

Related fix proposed to branch: master
Review: https://review.opendev.org/709181

[OpenStack Infra (hudson-openstack)]

Related fix proposed to branch: master
Review: https://review.opendev.org/709182

[OpenStack Infra (hudson-openstack)]

Reviewed: https://review.opendev.org/709181
Committed: https://git.openstack.org/cgit/openstack/tripleo-heat-templates/commit/?id=4d8eb351147c359a6a2dea4ee43b247c3be59da8
Submitter: Zuul
Branch: master

commit 4d8eb351147c359a6a2dea4ee43b247c3be59da8
Author: Michele Baldessari <email address hidden>
Date: Fri Feb 21 17:05:46 2020 +0100

    Drop bootstrap_host_exec from pacemaker_restart_bundle

    bootstrap_host_exec does not exist on the host so we cannot assume its
    existence. Since, barring argument checking, it is three lines of shell
    script [1] we just use it directly. We also add an extra echo to make it
    simpler to debug any bootstrap vs non-bootstrap issued.

    [1] https://github.com/openstack/tripleo-common/blob/master/scripts/bootstrap_host_exec

    Change-Id: Ia850286682f09cd75651591a1158c2e467343c1d
    Related-Bug: #1863442

[OpenStack Infra (hudson-openstack)]

Related fix proposed to branch: master
Review: https://review.opendev.org/724817

[OpenStack Infra (hudson-openstack)]

Change abandoned by Michele Baldessari (<email address hidden>) on branch: master
Review: https://review.opendev.org/709182

[OpenStack Infra (hudson-openstack)]

Reviewed: https://review.opendev.org/708738
Committed: https://git.openstack.org/cgit/openstack/puppet-tripleo/commit/?id=d185cbf032e02eec7f051e85c51c19732620e192
Submitter: Zuul
Branch: master

commit d185cbf032e02eec7f051e85c51c19732620e192
Author: Michele Baldessari <email address hidden>
Date: Wed Feb 19 21:07:09 2020 +0100

    Allow OCF resources to be created with --force

    While moving to running pcs commands on the host and off short-lived
    containers, we are confronted with the issue that pcs usually checks
    for the resource agent's existence on the host before creating it.
    Since we'd rather avoid installing the needed resource agents on the
    host (as it is inside a container), we allow a new 'force_ocf' parameter
    to be passed to those situations where we might need it.

    Depends-On: I20eb78a061a334b20f6b2274591c5d313a0af532

    Related-Bug: #1863442
    Change-Id: If9048196b5c03e3cfaba72f043b7f7275568bdc4

[OpenStack Infra (hudson-openstack)]

Related fix proposed to branch: stable/ussuri
Review: https://review.opendev.org/731061

[OpenStack Infra (hudson-openstack)]

Reviewed: https://review.opendev.org/731061
Committed: https://git.openstack.org/cgit/openstack/puppet-tripleo/commit/?id=844975ef4be5e52cd056b7a11a4251ae1ef53ef5
Submitter: Zuul
Branch: stable/ussuri

commit 844975ef4be5e52cd056b7a11a4251ae1ef53ef5
Author: Michele Baldessari <email address hidden>
Date: Wed Feb 19 21:07:09 2020 +0100

    Allow OCF resources to be created with --force

    While moving to running pcs commands on the host and off short-lived
    containers, we are confronted with the issue that pcs usually checks
    for the resource agent's existence on the host before creating it.
    Since we'd rather avoid installing the needed resource agents on the
    host (as it is inside a container), we allow a new 'force_ocf' parameter
    to be passed to those situations where we might need it.

    Depends-On: I20eb78a061a334b20f6b2274591c5d313a0af532

    Related-Bug: #1863442
    Change-Id: If9048196b5c03e3cfaba72f043b7f7275568bdc4
    (cherry picked from commit d185cbf032e02eec7f051e85c51c19732620e192)

[OpenStack Infra (hudson-openstack)]

Related fix proposed to branch: stable/train
Review: https://review.opendev.org/746937

[OpenStack Infra (hudson-openstack)]

Reviewed: https://review.opendev.org/746937
Committed: https://git.openstack.org/cgit/openstack/tripleo-heat-templates/commit/?id=da6b7b424c2eb709e22cf1d809bde405cf8e818c
Submitter: Zuul
Branch: stable/train

commit da6b7b424c2eb709e22cf1d809bde405cf8e818c
Author: Michele Baldessari <email address hidden>
Date: Fri Feb 21 17:05:46 2020 +0100

    Drop bootstrap_host_exec from pacemaker_restart_bundle

    bootstrap_host_exec does not exist on the host so we cannot assume its
    existence. Since, barring argument checking, it is three lines of shell
    script [1] we just use it directly. We also add an extra echo to make it
    simpler to debug any bootstrap vs non-bootstrap issued.

    [1] https://github.com/openstack/tripleo-common/blob/master/scripts/bootstrap_host_exec

    Change-Id: Ia850286682f09cd75651591a1158c2e467343c1d
    Related-Bug: #1863442
    (cherry picked from commit 4d8eb351147c359a6a2dea4ee43b247c3be59da8)

[OpenStack Infra (hudson-openstack)]

Reviewed: https://review.opendev.org/724817
Committed: https://git.openstack.org/cgit/openstack/tripleo-ansible/commit/?id=d5782863b8b1da748da3e6914ff4281e6103f29b
Submitter: Zuul
Branch: master

commit d5782863b8b1da748da3e6914ff4281e6103f29b
Author: Michele Baldessari <email address hidden>
Date: Fri May 1 12:43:46 2020 +0200

    Add tripleo_ha_wrapper role

    This role is meant to help moving the HA pcs tasks to pure deployment
    steps run on the host and not in containers any more. We tested this
    ansible review with the THT patches at
    https://review.opendev.org/#/q/status:open+project:openstack/tripleo-heat-templates+branch:master+topic:pcs_host3

    Here is the testing protocol that was used in order to verify we did
    not break anything:
    1) Deployed overcloud + tempest
    2) Ran a successful redeploy
    2.1) Verified that a redeploy restarted containers globally when the
         configuration for the service changed
    2.2) Verified that a redeploy did not restart containers which config
         did not change
    2.3) Verified that the ordering of restarting bundles allows for
         brownfield deployments (i.e. first the bundle is restarted due
         to new bind-mounts and then the bundle is restarted with new
         configuration)
    2.4) Verified that converge step after minor update is successful
    3) Ran an initial successful minor update
    3.1) Verified that the new image is indeed being run by pacemaker
         (i.e. :pcmklatest tag is applied to the new containers)
    3.2) Verified that after converge no container is restarted
    4) Re-run a subsequent successful noop minor update
    4.1) Verify that containers are not restarted a second time after
         the cluster is stopped and started

    Depends-On: I367cf4b65300be8dca0190b9adeab549018d4a56

    Change-Id: Iaa7e89f0d25221c2a6ef0b81eb88a6f496f01696
    Related-Bug: #1863442

[OpenStack Infra (hudson-openstack)]

Reviewed: https://review.opendev.org/707955
Committed: https://git.openstack.org/cgit/openstack/tripleo-heat-templates/commit/?id=da3d5e80562f7aa3d0039d44a2fd0a2f4c093cf4
Submitter: Zuul
Branch: master

commit da3d5e80562f7aa3d0039d44a2fd0a2f4c093cf4
Author: Michele Baldessari <email address hidden>
Date: Fri May 1 13:12:46 2020 +0200

    pcs commands on host: mysql

    This implements the creation of the haproxy bundle on the host.
    The testing protocol used is documented in the depends-on.

    The reason for adding a post_update task is that during a minor update
    the deployment tasks are not run during the node update procedure but
    only during the final converge. So we ran the role again there to make
    sure that any config change will trigger a restart during the minor
    update, so the disruption is only local to the single node being
    updated. If we did not do this a final converge could potentially
    trigger a global restart of HA bundles which would be quite disruptive.

    NB: in this case we keep the container init_bundle (renamed to
    wait_bundle) around just use it to wait for galera to be up.

    Depends-On: Iaa7e89f0d25221c2a6ef0b81eb88a6f496f01696

    Change-Id: Ie14819b66cecdb5a9cc6299b68a0cc70a7aa3370
    Related-Bug: #1863442

[OpenStack Infra (hudson-openstack)]

Reviewed: https://review.opendev.org/707953
Committed: https://git.openstack.org/cgit/openstack/tripleo-heat-templates/commit/?id=a11622359002597c7f32c6efeda745b8500a34e6
Submitter: Zuul
Branch: master

commit a11622359002597c7f32c6efeda745b8500a34e6
Author: Michele Baldessari <email address hidden>
Date: Fri May 1 13:07:13 2020 +0200

    pcs commands on host: haproxy bundle

    This implements the creation of the haproxy bundle on the host.
    The testing protocol used is documented in the depends-on.

    The reason for adding a post_update task is that during a minor update
    the deployment tasks are not run during the node update procedure but
    only during the final converge. So we ran the role again there to make
    sure that any config change will trigger a restart during the minor
    update, so the disruption is only local to the single node being
    updated. If we did not do this a final converge could potentially
    trigger a global restart of HA bundles which would be quite disruptive.

    NB: We had to remove a few tags that were present befaure because we want
    to avoid that puppet checks for /etc/haproxy/haproxy.cfg existence
    on the host.

    Related-Bug: #1863442

    Depends-On: Iaa7e89f0d25221c2a6ef0b81eb88a6f496f01696

    Change-Id: I20dc57a02a315ca361883b73e4fb1cb98cf438cb

[OpenStack Infra (hudson-openstack)]

Reviewed: https://review.opendev.org/708357
Committed: https://git.openstack.org/cgit/openstack/tripleo-heat-templates/commit/?id=a1bee391361cde4345c2b78770865dab0278785e
Submitter: Zuul
Branch: master

commit a1bee391361cde4345c2b78770865dab0278785e
Author: Michele Baldessari <email address hidden>
Date: Fri May 1 13:28:36 2020 +0200

    pcs commands on host: manila-share

    This implements the creation of the manila-share bundle on the host.
    The testing protocol used is documented in the depends-on.

    The reason for adding a post_update task is that during a minor update
    the deployment tasks are not run during the node update procedure but
    only during the final converge. So we ran the role again there to make
    sure that any config change will trigger a restart during the minor
    update, so the disruption is only local to the single node being
    updated. If we did not do this a final converge could potentially
    trigger a global restart of HA bundles which would be quite disruptive.

    Depends-On: Iaa7e89f0d25221c2a6ef0b81eb88a6f496f01696

    Change-Id: If5730dba4973a5927c84565b6a65398ea1d7072f
    Related-Bug: #1863442

[OpenStack Infra (hudson-openstack)]

Reviewed: https://review.opendev.org/707956
Committed: https://git.openstack.org/cgit/openstack/tripleo-heat-templates/commit/?id=f37f21f31df12e553d1940fce08195fcb18ccf3c
Submitter: Zuul
Branch: master

commit f37f21f31df12e553d1940fce08195fcb18ccf3c
Author: Michele Baldessari <email address hidden>
Date: Fri May 1 13:19:58 2020 +0200

    pcs commands on host: cinder backup/volume

    This implements the creation of the haproxy bundle on the host.
    The testing protocol used is documented in the depends-on.

    The reason for adding a post_update task is that during a minor update
    the deployment tasks are not run during the node update procedure but
    only during the final converge. So we ran the role again there to make
    sure that any config change will trigger a restart during the minor
    update, so the disruption is only local to the single node being
    updated. If we did not do this a final converge could potentially
    trigger a global restart of HA bundles which would be quite disruptive.

    Depends-On: Iaa7e89f0d25221c2a6ef0b81eb88a6f496f01696
    Change-Id: Ia4399b632257e693fb2c516e487856331149589d
    Related-Bug: #1863442

[OpenStack Infra (hudson-openstack)]

Reviewed: https://review.opendev.org/707954
Committed: https://git.openstack.org/cgit/openstack/tripleo-heat-templates/commit/?id=83d1cb585268177ef6f685e689d5aa34b24c8056
Submitter: Zuul
Branch: master

commit 83d1cb585268177ef6f685e689d5aa34b24c8056
Author: Michele Baldessari <email address hidden>
Date: Tue Aug 11 15:04:36 2020 +0200

    pcs commands on host: redis bundle

    This implements the creation of the redis bundle on the host.
    The testing protocol used is documented in the depends-on.

    The full rationale is contained in the LP bug.

    The reason for adding a post_update task is that during a minor update
    the deployment tasks are not run during the node update procedure but
    only during the final converge. So we ran the role again there to make
    sure that any config change will trigger a restart during the minor
    update, so the disruption is only local to the single node being
    updated. If we did not do this a final converge could potentially
    trigger a global restart of HA bundles which would be quite disruptive.

    Related-Bug: #1863442

    Depends-On: Iaa7e89f0d25221c2a6ef0b81eb88a6f496f01696

    Change-Id: I5ce8367363d535b71b01395b0bef4cf17c8935b5

[OpenStack Infra (hudson-openstack)]

Reviewed: https://review.opendev.org/707957
Committed: https://git.openstack.org/cgit/openstack/tripleo-heat-templates/commit/?id=4e76e1fb72ce1af012abf412c179d286fd127bd9
Submitter: Zuul
Branch: master

commit 4e76e1fb72ce1af012abf412c179d286fd127bd9
Author: Michele Baldessari <email address hidden>
Date: Fri May 1 13:23:46 2020 +0200

    pcs commands on host: ovn dbs

    We add the following depends on because at this stage ovn will be
    created on the host and the resource-agent won't be present so
    we have to be able to tell pcs to ignore the non existance of the
    ovn-dbs ocfg resource on the host.

    Depends-On: If9048196b5c03e3cfaba72f043b7f7275568bdc4
    Depends-On: Iaa7e89f0d25221c2a6ef0b81eb88a6f496f01696

    Change-Id: I918b6c16db6ed70d9ad612aecd7af7d725520f7b
    Related-Bug: #1863442

[OpenStack Infra (hudson-openstack)]

Related fix proposed to branch: stable/ussuri
Review: https://review.opendev.org/750194

[OpenStack Infra (hudson-openstack)]

Reviewed: https://review.opendev.org/750194
Committed: https://git.openstack.org/cgit/openstack/tripleo-ansible/commit/?id=8daba50a383d303f8b991ca1b3c79c46cf73a85b
Submitter: Zuul
Branch: stable/ussuri

commit 8daba50a383d303f8b991ca1b3c79c46cf73a85b
Author: Michele Baldessari <email address hidden>
Date: Fri May 1 12:43:46 2020 +0200

    Add tripleo_ha_wrapper role

    This role is meant to help moving the HA pcs tasks to pure deployment
    steps run on the host and not in containers any more. We tested this
    ansible review with the THT patches at
    https://review.opendev.org/#/q/status:open+project:openstack/tripleo-heat-templates+branch:master+topic:pcs_host3

    Here is the testing protocol that was used in order to verify we did
    not break anything:
    1) Deployed overcloud + tempest
    2) Ran a successful redeploy
    2.1) Verified that a redeploy restarted containers globally when the
         configuration for the service changed
    2.2) Verified that a redeploy did not restart containers which config
         did not change
    2.3) Verified that the ordering of restarting bundles allows for
         brownfield deployments (i.e. first the bundle is restarted due
         to new bind-mounts and then the bundle is restarted with new
         configuration)
    2.4) Verified that converge step after minor update is successful
    3) Ran an initial successful minor update
    3.1) Verified that the new image is indeed being run by pacemaker
         (i.e. :pcmklatest tag is applied to the new containers)
    3.2) Verified that after converge no container is restarted
    4) Re-run a subsequent successful noop minor update
    4.1) Verify that containers are not restarted a second time after
         the cluster is stopped and started

    Depends-On: I367cf4b65300be8dca0190b9adeab549018d4a56

    Change-Id: Iaa7e89f0d25221c2a6ef0b81eb88a6f496f01696
    Related-Bug: #1863442
    (cherry picked from commit d5782863b8b1da748da3e6914ff4281e6103f29b)

[OpenStack Infra (hudson-openstack)]

Related fix proposed to branch: stable/train
Review: https://review.opendev.org/750277

[OpenStack Infra (hudson-openstack)]

Reviewed: https://review.opendev.org/750277
Committed: https://git.openstack.org/cgit/openstack/tripleo-ansible/commit/?id=53ccc3fddf97d6efc226008f8bc52da2ebba90fc
Submitter: Zuul
Branch: stable/train

commit 53ccc3fddf97d6efc226008f8bc52da2ebba90fc
Author: Michele Baldessari <email address hidden>
Date: Fri May 1 12:43:46 2020 +0200

    Add tripleo_ha_wrapper role

    This role is meant to help moving the HA pcs tasks to pure deployment
    steps run on the host and not in containers any more. We tested this
    ansible review with the THT patches at
    https://review.opendev.org/#/q/status:open+project:openstack/tripleo-heat-templates+branch:master+topic:pcs_host3

    Here is the testing protocol that was used in order to verify we did
    not break anything:
    1) Deployed overcloud + tempest
    2) Ran a successful redeploy
    2.1) Verified that a redeploy restarted containers globally when the
         configuration for the service changed
    2.2) Verified that a redeploy did not restart containers which config
         did not change
    2.3) Verified that the ordering of restarting bundles allows for
         brownfield deployments (i.e. first the bundle is restarted due
         to new bind-mounts and then the bundle is restarted with new
         configuration)
    2.4) Verified that converge step after minor update is successful
    3) Ran an initial successful minor update
    3.1) Verified that the new image is indeed being run by pacemaker
         (i.e. :pcmklatest tag is applied to the new containers)
    3.2) Verified that after converge no container is restarted
    4) Re-run a subsequent successful noop minor update
    4.1) Verify that containers are not restarted a second time after
         the cluster is stopped and started

    Depends-On: I367cf4b65300be8dca0190b9adeab549018d4a56

    NB: Not 100% clean in zuul.d/molecule.yaml
        Small tweaks/renames to molecule folder for CI

    Change-Id: Iaa7e89f0d25221c2a6ef0b81eb88a6f496f01696
    Related-Bug: #1863442
    (cherry picked from commit d5782863b8b1da748da3e6914ff4281e6103f29b)
    (cherry picked from commit 8daba50a383d303f8b991ca1b3c79c46cf73a85b)

[OpenStack Infra (hudson-openstack)]

Related fix proposed to branch: stable/ussuri
Review: https://review.opendev.org/750455

[OpenStack Infra (hudson-openstack)]

Related fix proposed to branch: stable/ussuri
Review: https://review.opendev.org/750456

[OpenStack Infra (hudson-openstack)]

Related fix proposed to branch: stable/ussuri
Review: https://review.opendev.org/750457

[OpenStack Infra (hudson-openstack)]

Related fix proposed to branch: stable/ussuri
Review: https://review.opendev.org/750458

[OpenStack Infra (hudson-openstack)]

Related fix proposed to branch: stable/ussuri
Review: https://review.opendev.org/750459

[OpenStack Infra (hudson-openstack)]

Related fix proposed to branch: stable/ussuri
Review: https://review.opendev.org/750460

[OpenStack Infra (hudson-openstack)]

Change abandoned by Michele Baldessari (<email address hidden>) on branch: stable/ussuri
Review: https://review.opendev.org/750455
Reason: don't want to merge this before we understand the upgrade failures on the other pcs host patches

[OpenStack Infra (hudson-openstack)]

Related fix proposed to branch: master
Review: https://review.opendev.org/750678

[OpenStack Infra (hudson-openstack)]

Related fix proposed to branch: master
Review: https://review.opendev.org/750763

[OpenStack Infra (hudson-openstack)]

Reviewed: https://review.opendev.org/750763
Committed: https://git.openstack.org/cgit/openstack/tripleo-heat-templates/commit/?id=a584099b9d581f48011e42acb3ac7f173f39b479
Submitter: Zuul
Branch: master

commit a584099b9d581f48011e42acb3ac7f173f39b479
Author: Michele Baldessari <email address hidden>
Date: Wed Sep 9 19:48:14 2020 +0200

    Make sure we noop the Rabbitmq_ providers

    This is needed because the Rabbitmq_ providers might prefetch
    the data from rabbitmq on the system and we do not want to run
    any task on the host that is not strictly pcmk-related.

    Co-Authored-By: Damien Ciabrini <email address hidden>

    Change-Id: I0762b3ed7d881babfecc49b1a0e3798eb3efbddb
    Related-Bug: #1863442

[OpenStack Infra (hudson-openstack)]

Reviewed: https://review.opendev.org/750678
Committed: https://git.openstack.org/cgit/openstack/tripleo-heat-templates/commit/?id=4b0cc36efe9265987be39359070c5f70260af8ff
Submitter: Zuul
Branch: master

commit 4b0cc36efe9265987be39359070c5f70260af8ff
Author: Michele Baldessari <email address hidden>
Date: Wed Sep 9 16:37:18 2020 +0200

    Make sure we noop the Mysql_ providers

    This is needed because the Mysql_ providers will prefetch
    the the mysql users if facter finds the '/bin/mysql' executable
    on the system and we do not want to run any mysql task on the host
    directly.

    Co-Authored-By: Damien Ciabrini <email address hidden>

    Change-Id: Ic6c65e6849368185177aeaa31d50f52761225f62
    Related-Bug: #1863442

[OpenStack Infra (hudson-openstack)]

Reviewed: https://review.opendev.org/750455
Committed: https://git.openstack.org/cgit/openstack/tripleo-heat-templates/commit/?id=3f8c22b5614253d201b834f4d1586da4a386391d
Submitter: Zuul
Branch: stable/ussuri

commit 3f8c22b5614253d201b834f4d1586da4a386391d
Author: Michele Baldessari <email address hidden>
Date: Fri May 1 13:07:13 2020 +0200

    pcs commands on host: haproxy bundle

    This implements the creation of the haproxy bundle on the host.
    The testing protocol used is documented in the depends-on.

    The reason for adding a post_update task is that during a minor update
    the deployment tasks are not run during the node update procedure but
    only during the final converge. So we ran the role again there to make
    sure that any config change will trigger a restart during the minor
    update, so the disruption is only local to the single node being
    updated. If we did not do this a final converge could potentially
    trigger a global restart of HA bundles which would be quite disruptive.

    NB: We had to remove a few tags that were present befaure because we want
    to avoid that puppet checks for /etc/haproxy/haproxy.cfg existence
    on the host.

    Related-Bug: #1863442

    Depends-On: Iaa7e89f0d25221c2a6ef0b81eb88a6f496f01696

    Change-Id: I20dc57a02a315ca361883b73e4fb1cb98cf438cb
    (cherry picked from commit a11622359002597c7f32c6efeda745b8500a34e6)

[OpenStack Infra (hudson-openstack)]

Reviewed: https://review.opendev.org/750456
Committed: https://git.openstack.org/cgit/openstack/tripleo-heat-templates/commit/?id=be1fee4af6a5ed7ef4eff4ede26f5d3666132398
Submitter: Zuul
Branch: stable/ussuri

commit be1fee4af6a5ed7ef4eff4ede26f5d3666132398
Author: Michele Baldessari <email address hidden>
Date: Fri May 1 13:12:46 2020 +0200

    pcs commands on host: mysql

    This implements the creation of the haproxy bundle on the host.
    The testing protocol used is documented in the depends-on.

    The reason for adding a post_update task is that during a minor update
    the deployment tasks are not run during the node update procedure but
    only during the final converge. So we ran the role again there to make
    sure that any config change will trigger a restart during the minor
    update, so the disruption is only local to the single node being
    updated. If we did not do this a final converge could potentially
    trigger a global restart of HA bundles which would be quite disruptive.

    NB: in this case we keep the container init_bundle (renamed to
    wait_bundle) around just use it to wait for galera to be up.

    Depends-On: Iaa7e89f0d25221c2a6ef0b81eb88a6f496f01696

    NB: We also merged Ic6c65e6849368185177aeaa31d50f52761225f62
        in this review so that the ussuri upgrade job can pass
        (Make sure we noop the Mysql_ providers)

    Change-Id: Ie14819b66cecdb5a9cc6299b68a0cc70a7aa3370
    Related-Bug: #1863442
    (cherry picked from commit da3d5e80562f7aa3d0039d44a2fd0a2f4c093cf4)

[OpenStack Infra (hudson-openstack)]

Reviewed: https://review.opendev.org/750457
Committed: https://git.openstack.org/cgit/openstack/tripleo-heat-templates/commit/?id=06209fa28f35f968235fc96989a333fb950df850
Submitter: Zuul
Branch: stable/ussuri

commit 06209fa28f35f968235fc96989a333fb950df850
Author: Michele Baldessari <email address hidden>
Date: Fri May 1 13:28:36 2020 +0200

    pcs commands on host: manila-share

    This implements the creation of the manila-share bundle on the host.
    The testing protocol used is documented in the depends-on.

    The reason for adding a post_update task is that during a minor update
    the deployment tasks are not run during the node update procedure but
    only during the final converge. So we ran the role again there to make
    sure that any config change will trigger a restart during the minor
    update, so the disruption is only local to the single node being
    updated. If we did not do this a final converge could potentially
    trigger a global restart of HA bundles which would be quite disruptive.

    Depends-On: Iaa7e89f0d25221c2a6ef0b81eb88a6f496f01696

    Change-Id: If5730dba4973a5927c84565b6a65398ea1d7072f
    Related-Bug: #1863442
    (cherry picked from commit a1bee391361cde4345c2b78770865dab0278785e)

[OpenStack Infra (hudson-openstack)]

Reviewed: https://review.opendev.org/750458
Committed: https://git.openstack.org/cgit/openstack/tripleo-heat-templates/commit/?id=4f56371f912f468758c3533079872b7e69d4c32f
Submitter: Zuul
Branch: stable/ussuri

commit 4f56371f912f468758c3533079872b7e69d4c32f
Author: Michele Baldessari <email address hidden>
Date: Fri May 1 13:19:58 2020 +0200

    pcs commands on host: cinder backup/volume

    This implements the creation of the haproxy bundle on the host.
    The testing protocol used is documented in the depends-on.

    The reason for adding a post_update task is that during a minor update
    the deployment tasks are not run during the node update procedure but
    only during the final converge. So we ran the role again there to make
    sure that any config change will trigger a restart during the minor
    update, so the disruption is only local to the single node being
    updated. If we did not do this a final converge could potentially
    trigger a global restart of HA bundles which would be quite disruptive.

    Depends-On: Iaa7e89f0d25221c2a6ef0b81eb88a6f496f01696
    Change-Id: Ia4399b632257e693fb2c516e487856331149589d
    Related-Bug: #1863442
    (cherry picked from commit f37f21f31df12e553d1940fce08195fcb18ccf3c)

[OpenStack Infra (hudson-openstack)]

Reviewed: https://review.opendev.org/750459
Committed: https://git.openstack.org/cgit/openstack/tripleo-heat-templates/commit/?id=ec70716ad874d36babbefe8852c1e9a37f1518b8
Submitter: Zuul
Branch: stable/ussuri

commit ec70716ad874d36babbefe8852c1e9a37f1518b8
Author: Michele Baldessari <email address hidden>
Date: Fri May 1 13:23:46 2020 +0200

    pcs commands on host: ovn dbs

    We add the following depends on because at this stage ovn will be
    created on the host and the resource-agent won't be present so
    we have to be able to tell pcs to ignore the non existance of the
    ovn-dbs ocfg resource on the host.

    Depends-On: If9048196b5c03e3cfaba72f043b7f7275568bdc4
    Depends-On: Iaa7e89f0d25221c2a6ef0b81eb88a6f496f01696

    Change-Id: I918b6c16db6ed70d9ad612aecd7af7d725520f7b
    Related-Bug: #1863442
    (cherry picked from commit 4e76e1fb72ce1af012abf412c179d286fd127bd9)

[OpenStack Infra (hudson-openstack)]

Reviewed: https://review.opendev.org/750460
Committed: https://git.openstack.org/cgit/openstack/tripleo-heat-templates/commit/?id=0eb998930011ee3fc13ecc188dbeb578f185b1ec
Submitter: Zuul
Branch: stable/ussuri

commit 0eb998930011ee3fc13ecc188dbeb578f185b1ec
Author: Michele Baldessari <email address hidden>
Date: Tue Aug 11 15:04:36 2020 +0200

    pcs commands on host: redis bundle

    This implements the creation of the redis bundle on the host.
    The testing protocol used is documented in the depends-on.

    The full rationale is contained in the LP bug.

    The reason for adding a post_update task is that during a minor update
    the deployment tasks are not run during the node update procedure but
    only during the final converge. So we ran the role again there to make
    sure that any config change will trigger a restart during the minor
    update, so the disruption is only local to the single node being
    updated. If we did not do this a final converge could potentially
    trigger a global restart of HA bundles which would be quite disruptive.

    Related-Bug: #1863442

    Depends-On: Iaa7e89f0d25221c2a6ef0b81eb88a6f496f01696

    Change-Id: I5ce8367363d535b71b01395b0bef4cf17c8935b5
    (cherry picked from commit 83d1cb585268177ef6f685e689d5aa34b24c8056)

[OpenStack Infra (hudson-openstack)]

Reviewed: https://review.opendev.org/750461
Committed: https://git.openstack.org/cgit/openstack/tripleo-heat-templates/commit/?id=0ad2358067c0aff0bd36c1b31c8d9eaba9b3280c
Submitter: Zuul
Branch: stable/ussuri

commit 0ad2358067c0aff0bd36c1b31c8d9eaba9b3280c
Author: Michele Baldessari <email address hidden>
Date: Fri May 1 13:42:26 2020 +0200

    pcs commands on host: rabbitmq

    This implements the creation of the rabbitmq bundle on the host.
    The testing protocol used is documented in the depends-on.

    The reason for adding a post_update task is that during a minor update
    the deployment tasks are not run during the node update procedure but
    only during the final converge. So we ran the role again there to make
    sure that any config change will trigger a restart during the minor
    update, so the disruption is only local to the single node being
    updated. If we did not do this a final converge could potentially
    trigger a global restart of HA bundles which would be quite disruptive.

    NB: The init_bundle now has become the wait_bundle
    as it just waits for rabbitmq to be up and functional

    NB2: Slightly not 100% due to different context (missing
    I1873baab4325521dbcdc142462940ee3cf370c1b)

    NB3: We cherry-picked I0762b3ed7d881babfecc49b1a0e3798eb3efbddb
         inside this review in order for the upgrade job to pass
         (Make sure we noop the Rabbitmq_ providers)

    Related-Bug: #1863442

    Depends-On: Iaa7e89f0d25221c2a6ef0b81eb88a6f496f01696
    Change-Id: I853bcf354f64ef88cec9e68ad5c123e4af786de3
    (cherry picked from commit 29bc3de01dc39bb02eb7e7c27cc3661c591143bb)

[Marios Andreou (marios-b)]

This is an automated action. Bug status has been set to 'Incomplete' and target milestone has been removed due to inactivity. If you disagree please re-set these values and reach out to us on freenode #tripleo

[Launchpad Janitor (janitor)]

[Expired for tripleo because there has been no activity for 60 days.]