crm_node -n needlessly called via facter

[Michele Baldessari]

So in lib/facter/pacemaker_node_name.rb we do the following:
Facter.add('pacemaker_node_name') do
  setcode do
    Facter::Core::Execution.exec 'crm_node -n'
  end
end

This is problematic because starting with pacemaker 1.1.19 crm_node -n will trigger
a newly-introduced CRM_OP_NODE_INFO query which will hang if the container runs 1.1.19 code and the cluster runs 1.1.18 code.

Let's simply avoid running these queries when run inside a container

[OpenStack Infra (hudson-openstack)]

Fix proposed to branch: master
Review: https://review.openstack.org/583458

[OpenStack Infra (hudson-openstack)]

Reviewed: https://review.openstack.org/583458
Committed: https://git.openstack.org/cgit/openstack/puppet-pacemaker/commit/?id=966943a79e11eccac5ec0868836239c55e3845dc
Submitter: Zuul
Branch: master

commit 966943a79e11eccac5ec0868836239c55e3845dc
Author: Michele Baldessari <email address hidden>
Date: Wed Jul 18 07:44:00 2018 +0200

    Do not call crm_node -n when running in a container

    In lib/facter/pacemaker_node_name.rb we do the following:
    Facter.add('pacemaker_node_name') do
      setcode do
        Facter::Core::Execution.exec 'crm_node -n'
      end
    end

    This is problematic because starting with pacemaker 1.1.19 crm_node -n
    will trigger a newly-introduced CRM_OP_NODE_INFO query which will hang
    if the container runs 1.1.19 code and the cluster runs 1.1.18 code.

    Let's simply avoid running these queries when run inside a container.

    Tested this by deploying successfully an overcloud with pcmk-1.1.19 in
    containers and 1.1.18 on the host:
    [root@controller-0 ~]# rpm -q pacemaker
    pacemaker-1.1.18-11.el7_5.3.x86_64
    [root@controller-0 ~]# docker exec -it galera-bundle-docker-0 sh -c "rpm -q pacemaker"
    pacemaker-1.1.19-3.el7.x86_64

    For this to work we also need the following RA fix:
    https://github.com/ClusterLabs/resource-agents/pull/1173/

    Previously this would fail with crm_node -n just hanging.

    Change-Id: I9ae7df2f49f918507c5f98b2441b5b17423a38da
    Closes-Bug: #1782231

[OpenStack Infra (hudson-openstack)]

Fix proposed to branch: stable/0.6.x
Review: https://review.openstack.org/583660

[OpenStack Infra (hudson-openstack)]

Reviewed: https://review.openstack.org/583660
Committed: https://git.openstack.org/cgit/openstack/puppet-pacemaker/commit/?id=5a6dc500ae6f4c89466dad5436f341999f1ac38a
Submitter: Zuul
Branch: stable/0.6.x

commit 5a6dc500ae6f4c89466dad5436f341999f1ac38a
Author: Michele Baldessari <email address hidden>
Date: Wed Jul 18 07:44:00 2018 +0200

    Do not call crm_node -n when running in a container

    In lib/facter/pacemaker_node_name.rb we do the following:
    Facter.add('pacemaker_node_name') do
      setcode do
        Facter::Core::Execution.exec 'crm_node -n'
      end
    end

    This is problematic because starting with pacemaker 1.1.19 crm_node -n
    will trigger a newly-introduced CRM_OP_NODE_INFO query which will hang
    if the container runs 1.1.19 code and the cluster runs 1.1.18 code.

    Let's simply avoid running these queries when run inside a container.

    Tested this by deploying successfully an overcloud with pcmk-1.1.19 in
    containers and 1.1.18 on the host:
    [root@controller-0 ~]# rpm -q pacemaker
    pacemaker-1.1.18-11.el7_5.3.x86_64
    [root@controller-0 ~]# docker exec -it galera-bundle-docker-0 sh -c "rpm -q pacemaker"
    pacemaker-1.1.19-3.el7.x86_64

    For this to work we also need the following RA fix:
    https://github.com/ClusterLabs/resource-agents/pull/1173/

    Previously this would fail with crm_node -n just hanging.

    Change-Id: I9ae7df2f49f918507c5f98b2441b5b17423a38da
    Closes-Bug: #1782231
    (cherry picked from commit 966943a79e11eccac5ec0868836239c55e3845dc)

[OpenStack Infra (hudson-openstack)]

This issue was fixed in the openstack/puppet-pacemaker 0.7.2 release.