Gnocchi freezes if coordination URL is unresponsive

[Nicolas Vila]

Using gnocchi 2.1.1.dev36, when configuring a coordination URL that is unresponsive, gnocchi-metricd freezes upon start. When debug logging is enabled, metricd outputs all configuration parameters and then just stays there:

(...)
2016-05-20 10:05:07.831 12240 DEBUG gnocchi.service [-] api.workers = 2 log_opt_values /usr/local/lib/python2.7/dist-packages/oslo_config/cfg.py:2525
2016-05-20 10:05:07.831 12240 DEBUG gnocchi.service [-] archive_policy.default_aggregation_methods = ['mean', 'min', 'max', 'sum', 'std', 'median', 'count', '95pct'] log_opt_values /usr/local/lib/python2.7/dist-packages/oslo_config/cfg.py:2525
2016-05-20 10:05:07.831 12240 DEBUG gnocchi.service [-] oslo_policy.policy_default_rule = default log_opt_values /usr/local/lib/python2.7/dist-packages/oslo_config/cfg.py:2525
2016-05-20 10:05:07.832 12240 DEBUG gnocchi.service [-] oslo_policy.policy_dirs = ['policy.d'] log_opt_values /usr/local/lib/python2.7/dist-packages/oslo_config/cfg.py:2525
2016-05-20 10:05:07.832 12240 DEBUG gnocchi.service [-] oslo_policy.policy_file = policy.json log_opt_values /usr/local/lib/python2.7/dist-packages/oslo_config/cfg.py:2525
2016-05-20 10:05:07.832 12240 DEBUG gnocchi.service [-] ******************************************************************************** log_opt_values /usr/local/lib/python2.7/dist-packages/oslo_config/cfg.py:2527

It would be useful for metricd to log a message saying that the coordination URL is not working, it'd help a lot with troubleshooting.

Thanks, regards.

[OpenStack Infra (hudson-openstack)]

Fix proposed to branch: master
Review: https://review.openstack.org/322834

[OpenStack Infra (hudson-openstack)]

Fix proposed to branch: stable/2.1
Review: https://review.openstack.org/323407

[OpenStack Infra (hudson-openstack)]

Reviewed: https://review.openstack.org/322834
Committed: https://git.openstack.org/cgit/openstack/gnocchi/commit/?id=8932aad1c3fa95c9ee901aab2b5c2f089a020312
Submitter: Jenkins
Branch: master

commit 8932aad1c3fa95c9ee901aab2b5c2f089a020312
Author: Julien Danjou <email address hidden>
Date: Mon May 30 15:53:18 2016 +0200

    metricd: only retry on attended errors and print error when coordinator fails

    Currently, `retrying' retry on every error raised by either the storage or
    indexer. Explicitely retry on what is expected to be failing, and not on random
    errors. Also print a log indicating something is wrong.

    Change-Id: I7d3e94cf94a22dcc08fb529537f10758db7d58ca
    Closes-Bug: #1584083

[OpenStack Infra (hudson-openstack)]

Reviewed: https://review.openstack.org/323407
Committed: https://git.openstack.org/cgit/openstack/gnocchi/commit/?id=6c49492db68e921b515c819463c58e4800ea0dc8
Submitter: Jenkins
Branch: stable/2.1

commit 6c49492db68e921b515c819463c58e4800ea0dc8
Author: Julien Danjou <email address hidden>
Date: Mon May 30 15:53:18 2016 +0200

    metricd: only retry on attended errors and print error when coordinator fails

    Currently, `retrying' retry on every error raised by either the storage or
    indexer. Explicitely retry on what is expected to be failing, and not on random
    errors. Also print a log indicating something is wrong.

    Change-Id: I7d3e94cf94a22dcc08fb529537f10758db7d58ca
    Closes-Bug: #1584083
    (cherry picked from commit 8932aad1c3fa95c9ee901aab2b5c2f089a020312)

[Doug Hellmann (doug-hellmann)]

This issue was fixed in the openstack/gnocchi 2.2.0 release.