Ceilometer

  1. Bug #1336755
  2. Comment #10

Comment 10 for bug 1336755

Revision history for this message
Salvatore Orlando (salvatore-orlando) wrote :

This is currently the top cause of failures in the neutron full job.
And the neutron full job is where this error occurs more frequently:
http://logstash.openstack.org/#eyJmaWVsZHMiOltdLCJzZWFyY2giOiJtZXNzYWdlOlwiU2FtcGxlIGZvciBtZXRyaWM6bWVtb3J5IHdpdGggcXVlcnk6XCIgQU5EIG1lc3NhZ2U6XCJoYXMgbm90IGJlZW4gYWRkZWQgdG8gdGhlIGRhdGFiYXNlIHdpdGhpblwiIEFORCB0YWdzOmNvbnNvbGUiLCJ0aW1lZnJhbWUiOiI4NjQwMDAiLCJncmFwaG1vZGUiOiJjb3VudCIsIm9mZnNldCI6MCwidGltZSI6eyJ1c2VyX2ludGVydmFsIjowfSwibW9kZSI6InNjb3JlIiwiYW5hbHl6ZV9maWVsZCI6ImJ1aWxkX25hbWUifQ==

The actual frequency is even worse if one considers the neutron full job runs only for neutron and tempest patches.
So I would support the hypothesis of something affecting collector performance.