Time information missing from deploy/update/upgrade Ansible logs

[Jiří Stránský]

Some time ago we started getting detailed info about the time progress of deployment when we started printing Heat events together with timestamps in our deploy/upgrade commands. After moving to Ansible we don't have a comparable time information in the output.

This is not a serious drawback but it is something that would be nice to have to ease debugging. When debugging difficult problems, we can use every straw to hold on to, and very often it is *when* things happen that give clues to uncover root causes.

[OpenStack Infra (hudson-openstack)]

Fix proposed to branch: master
Review: https://review.openstack.org/576185

[OpenStack Infra (hudson-openstack)]

Reviewed: https://review.openstack.org/576185
Committed: https://git.openstack.org/cgit/openstack/tripleo-common/commit/?id=5dac6edf85c71df2f2cf722ae41e8018701a51c1
Submitter: Zuul
Branch: master

commit 5dac6edf85c71df2f2cf722ae41e8018701a51c1
Author: Jiri Stransky <email address hidden>
Date: Mon Jun 18 17:27:07 2018 +0200

    Provide timestamps in Ansible output by default

    Add a new configurable `profile_tasks` argument for
    AnsiblePlaybookAction that controls output of profiling
    information. The default for the new option is True, as it's likely
    that this will be a universally useful default for deployment,
    updates, upgrades, etc.

    Also a `profile_tasks_limit` argument is added to control final output
    of slowest executing tasks in the playbook. Since we want to optimize
    for production scenarios and operators can hardly affect the run
    times, this argument defaults to 0.

    The combination of the above will give us time stamps for each task
    but no "what are the slowest tasks" by default.

    Both options are overridable by extra_env_variables, which might be
    useful in some cases, especially for the callback whitelist.

    Change-Id: If8abc8c2406e152004c7bce1d116bc8690ac1cf7
    Closes-Bug: #1777467

[OpenStack Infra (hudson-openstack)]

Fix proposed to branch: stable/queens
Review: https://review.openstack.org/577103

[OpenStack Infra (hudson-openstack)]

Reviewed: https://review.openstack.org/577103
Committed: https://git.openstack.org/cgit/openstack/tripleo-common/commit/?id=255831dc0da1d5460cc82973fa3496e76752bf9f
Submitter: Zuul
Branch: stable/queens

commit 255831dc0da1d5460cc82973fa3496e76752bf9f
Author: Jiri Stransky <email address hidden>
Date: Mon Jun 18 17:27:07 2018 +0200

    Provide timestamps in Ansible output by default

    Add a new configurable `profile_tasks` argument for
    AnsiblePlaybookAction that controls output of profiling
    information. The default for the new option is True, as it's likely
    that this will be a universally useful default for deployment,
    updates, upgrades, etc.

    Also a `profile_tasks_limit` argument is added to control final output
    of slowest executing tasks in the playbook. Since we want to optimize
    for production scenarios and operators can hardly affect the run
    times, this argument defaults to 0.

    The combination of the above will give us time stamps for each task
    but no "what are the slowest tasks" by default.

    Both options are overridable by extra_env_variables, which might be
    useful in some cases, especially for the callback whitelist.

    Change-Id: If8abc8c2406e152004c7bce1d116bc8690ac1cf7
    Closes-Bug: #1777467
    (cherry picked from commit 5dac6edf85c71df2f2cf722ae41e8018701a51c1)

[OpenStack Infra (hudson-openstack)]

This issue was fixed in the openstack/tripleo-common 8.6.3 release.

[OpenStack Infra (hudson-openstack)]

This issue was fixed in the openstack/tripleo-common 9.2.0 release.