This fails constantly, and it looks like a socket error, an blocked or already used port. >23:20:53 TEST-PASS | /var/folders/Jj/Jjc6CMB3GWKy9EgG9MiLS++++TM/-Tmp-/tmpLQGFs7.mozmill-tests/firefox/tests/functional/restartTests/testSoftwareUpdateAutoProxy/test2.js | test2.js::teardownModule >23:20:53 INFO Passed: 113 >23:20:53 INFO Failed: 0 >23:20:53 INFO Skipped: 8 >23:20:53 Sending results to 'http://mozauto.iriscouch.com/mozmill-release/' failed ([Errno 35] Resource temporarily unavailable).
Why is that infrastructure? Looks like a better handling is needed on the sending side. Can you please check what would be necessary to do to get around this problem?
A week has been passed by here. Why hasn't this bug gotten an update yet? Also we might want to leave it in Infra for now.
It didn't fail latter again, when I connected on the nodes to investigated I tried to access the http://mozauto.iriscouch.com/mozmill-release/ in browser and I've got a response so I retriggered the jobs and they didn't failed again, though I didn't noted the affected nodes it wasn't only one mac(ex 10.6.8) type but most of them.
Setting as WFM for now.
Looks like this is in fact a dupe of bug 948912 (or the other way around). It's still happening, and should be a high priority.