Comment 10 for bug 1867380

Reviewed: https://review.opendev.org/713845
Committed: https://git.openstack.org/cgit/openstack/nova/commit/?id=0718015f3fd2899720613bfef789f7023f112e30
Submitter: Zuul
Branch: stable/queens

commit 0718015f3fd2899720613bfef789f7023f112e30
Author: Lee Yarwood <email address hidden>
Date: Fri Mar 13 16:51:01 2020 +0000

    nova-live-migration: Wait for n-cpu services to come up after configuring Ceph

    Previously the ceph.sh script used during the nova-live-migration job
    would only grep for a `compute` process when checking if the services
    had been restarted. This check was bogus and would always return 0 as it
    would always match itself. For example:

    2020-03-13 21:06:47.682073 | primary | 2020-03-13 21:06:47.681 | root
    29529 0.0 0.0 4500 736 pts/0 S+ 21:06 0:00 /bin/sh -c ps
           aux | grep compute
    2020-03-13 21:06:47.683964 | primary | 2020-03-13 21:06:47.683 | root
    29531 0.0 0.0 14616 944 pts/0 S+ 21:06 0:00 grep compute

    Failures of this job were seen on the stable/pike branch where slower CI
    nodes appeared to struggle to allow Libvirt to report to n-cpu in time
    before Tempest was started. This in-turn caused instance build failures
    and the overall failure of the job.

    This change resolves this issue by switching to pgrep and ensuring
    n-cpu services are reported as fully up after a cold restart before
    starting the Tempest test run.

    Closes-Bug: 1867380
    Change-Id: Icd7ab2ca4ddbed92c7e883a63a23245920d961e7
    (cherry picked from commit e23c3c2c8df3843c5853c87ef684bd21c4af95d8)
    (cherry picked from commit 70447bca2f4f33c6872eaf94a2e4351bb257c22a)
    (cherry picked from commit 373c4ffde2053c7ff11bd38339b88d144cd442f2)
    (cherry picked from commit 63ed32ef49adcb6830ef3b5329a561542bddf656)