The n-cpu and c-vol issues are from this change to python-glanceclient on 7/11 to use requests:
https://github.com/openstack/python- glanceclient/commit/dbb242b776908ca50ed8557ebfe7cfcd879366c8 That lines up with what we're seeing in the g-api logs with swiftclient and bug 1295812, and the times line up with logstash for n-cpu/c-vol logs and the fact they both use python-glanceclient. ** Summary changed: - HTTPConnectionPool is full warning should be a debug message + glanceclient is not handling http connection pools properly with python-requests ** Also affects: python-glanceclient Importance: Undecided Status: New ** No longer affects: glance -- You received this bug notification because you are a member of Yahoo! Engineering Team, which is subscribed to Glance. https://bugs.launchpad.net/bugs/1341777 Title: glanceclient is not handling http connection pools properly with python-requests Status in Python client library for Glance: New Bug description: This: message:"HttpConnectionPool is full, discarding connection: 127.0.0.1" AND tags:"screen-g-api.txt" shows up nearly 420K times in 7 days in Jenkins runs, on 100% successful jobs, so it's not a very good warning if it's expected to happen this much. It should be debug level if it's this pervasive, unless it's masking something else, but the logstash trends don't suggest that. http://logstash.openstack.org/#eyJzZWFyY2giOiJtZXNzYWdlOlwiSHR0cENvbm5lY3Rpb25Qb29sIGlzIGZ1bGwsIGRpc2NhcmRpbmcgY29ubmVjdGlvbjogMTI3LjAuMC4xXCIgQU5EIHRhZ3M6XCJzY3JlZW4tZy1hcGkudHh0XCIiLCJmaWVsZHMiOltdLCJvZmZzZXQiOjAsInRpbWVmcmFtZSI6IjYwNDgwMCIsImdyYXBobW9kZSI6ImNvdW50IiwidGltZSI6eyJ1c2VyX2ludGVydmFsIjowfSwic3RhbXAiOjE0MDUzNzE0NTU4NTZ9 To manage notifications about this bug go to: https://bugs.launchpad.net/python-glanceclient/+bug/1341777/+subscriptions -- Mailing list: https://launchpad.net/~yahoo-eng-team Post to : yahoo-eng-team@lists.launchpad.net Unsubscribe : https://launchpad.net/~yahoo-eng-team More help : https://help.launchpad.net/ListHelp
