check-tempest-dsvm-full fails with "fatal: read error: Connection reset by peer"
Affects | Status | Importance | Assigned to | Milestone | |
---|---|---|---|---|---|
OpenStack Core Infrastructure |
Triaged
|
Medium
|
Unassigned |
Bug Description
check-tempest-
2014-02-06 09:53:59.322 | Started by user anonymous
2014-02-06 09:53:59.325 | [EnvInject] - Loading node environment variables.
2014-02-06 09:54:15.906 | Building remotely on devstack-
2014-02-06 09:54:46.343 | [check-
2014-02-06 09:54:47.796 | + [[ ! -e devstack-gate ]]
2014-02-06 09:54:47.796 | + git clone git://git.
2014-02-06 09:54:48.161 | Cloning into 'devstack-gate'...
2014-02-06 09:54:48.442 | fatal: read error: Connection reset by peer
2014-02-06 09:54:48.443 | fatal: early EOF
2014-02-06 09:54:48.443 | fatal: index-pack failed
2014-02-06 09:54:48.829 | Build step 'Execute shell' marked build as failure
2014-02-06 09:54:50.074 | [SCP] Connecting to static.
2014-02-06 09:54:53.396 | [SCP] No file(s) found: logs/**
2014-02-06 09:54:53.875 | [SCP] ‘logs/**’ doesn’t match anything, but ‘**’ does. Perhaps that’s what you mean?
2014-02-06 09:54:53.954 | [SCP] Copying console log.
2014-02-06 09:54:54.792 | [SCP] Trying to create /srv/static/
2014-02-06 09:54:54.817 | Finished: FAILURE
Changed in openstack-ci: | |
milestone: | icehouse → kilo |
This looks like a disk I/O issue, and seems to have struck in hpcloud-region-b (where we don't run very many tests). Someone should mine logstash for additional occurrences and see whether that's mostly where they're coming from.