Stein upgrade: failing on pacemaker scale up

[Jiří Stránský]

Looks like the scale-up logic is outdated here:

https://github.com/openstack/puppet-pacemaker/blob/658f4ee6d982169191531bd0daf6ef8e08f06ddc/manifests/corosync.pp#L223

It fails, i had to scale up manually via e.g. `pcs cluster node add controller-2 addr=172.17.1.11 --start --wait`.

The error message produced is:

        "<13>Jul 1 16:53:28 puppet-user: Notice: /Stage[main]/Pacemaker::Corosync/Exec[Adding Cluster node: controller-1 to Cluster tripleo_cluster]/returns: Error: Cannot specify '--wait' without specifying '-
-start'",
        "<13>Jul 1 16:53:28 puppet-user: Notice: /Stage[main]/Pacemaker::Corosync/Exec[Adding Cluster node: controller-1 to Cluster tripleo_cluster]/returns: Error: Errors have occurred, therefore pcs is unable
 to continue",
        "<13>Jul 1 16:53:28 puppet-user: Notice: /Stage[main]/Pacemaker::Corosync/Exec[Adding Cluster node: controller-1 to Cluster tripleo_cluster]/returns: No addresses specified for host 'controller-1', usin
g 'controller-1'",
        "<13>Jul 1 16:53:28 puppet-user: Error: '/sbin/pcs cluster node add controller-1 --wait' returned 1 instead of one of [0]",
        "<13>Jul 1 16:53:28 puppet-user: Error: /Stage[main]/Pacemaker::Corosync/Exec[Adding Cluster node: controller-1 to Cluster tripleo_cluster]/returns: change from 'notrun' to ['0'] failed: '/sbin/pcs clus
ter node add controller-1 --wait' returned 1 instead of one of [0]",

[Jiří Stránský (jistr)]

Fix posted at: https://review.opendev.org/#/c/671011/

[OpenStack Infra (hudson-openstack)]

Reviewed: https://review.opendev.org/671011
Committed: https://git.openstack.org/cgit/openstack/puppet-pacemaker/commit/?id=d0b394d8f64d779c4866386f11f2c4c48f299ce5
Submitter: Zuul
Branch: master

commit d0b394d8f64d779c4866386f11f2c4c48f299ce5
Author: Jiri Stransky <email address hidden>
Date: Tue Jul 16 13:27:56 2019 +0200

    Scale up logic with extra IP addresses for pcs 0.10

    Nowadays we can provide IP addresses when adding a host into a
    pacemaker cluster. The pcmk_cluster_setup function for initial setup
    of the cluster has this already implemented, but it was still missing
    from the pcmk_nodes_added function and the scale-up logic. Support for
    scaling up with pcs 0.10 and the explicitly provided IP addresses is
    being added in this commit.

    Change-Id: If5baef83c739c7ccdbe346c75401f232ab092d20
    Closes-Bug: #1835217
    Related-Bug: #1831690

[OpenStack Infra (hudson-openstack)]

Fix proposed to branch: stable/0.7.x
Review: https://review.opendev.org/671718

[OpenStack Infra (hudson-openstack)]

Reviewed: https://review.opendev.org/671718
Committed: https://git.openstack.org/cgit/openstack/puppet-pacemaker/commit/?id=107e02f0b21131fcf3b42ec59c271fbd80f37b0f
Submitter: Zuul
Branch: stable/0.7.x

commit 107e02f0b21131fcf3b42ec59c271fbd80f37b0f
Author: Jiri Stransky <email address hidden>
Date: Tue Jul 16 13:27:56 2019 +0200

    Scale up logic with extra IP addresses for pcs 0.10

    Nowadays we can provide IP addresses when adding a host into a
    pacemaker cluster. The pcmk_cluster_setup function for initial setup
    of the cluster has this already implemented, but it was still missing
    from the pcmk_nodes_added function and the scale-up logic. Support for
    scaling up with pcs 0.10 and the explicitly provided IP addresses is
    being added in this commit.

    Change-Id: If5baef83c739c7ccdbe346c75401f232ab092d20
    Closes-Bug: #1835217
    Related-Bug: #1831690
    (cherry picked from commit d0b394d8f64d779c4866386f11f2c4c48f299ce5)

[OpenStack Infra (hudson-openstack)]

This issue was fixed in the openstack/puppet-pacemaker 0.8.0 release.