566889 - Some of the new batch of rev3 Fedora 64 slaves are not connecting

Status: RESOLVED FIXED

(Release Engineering :: General, defect, P3)

Description

[Armen [:armenzg]]

There is few slaves that cannot connect to the talos-master02 and it gets fixed by restarting the slave. I am not sure how long after being restarted this happens again and if it does happen again.



From looking at one of the logs and poking around:

The buildbot.tac seems to be correct.
The log ends saying that it got attached but the master says that it is disconnected.
Before this last message it had "lost remote" several times and the last time that it was running any steps was around 18:32PDT (24 mins before stop retrying).
"buildbot restart talos-slave" does not fix things.
Restarting the slave fixes the problem.
I have noticed that this machines have a date in the future.

2010/05/17 18:55 PDT [Broker,client] lost remote
2010/05/17 18:55 PDT [Broker,client] <twisted.internet.tcp.Connector instance at 0x1cd2050> will retry in 2 seconds
2010/05/17 18:55 PDT [Broker,client] Stopping factory <buildbot.slave.bot.BotFactory instance at 0x1e4fab8>
2010/05/17 18:55 PDT [-] Starting factory <buildbot.slave.bot.BotFactory instance at 0x1e4fab8>
2010/05/17 18:56 PDT [Broker,client] message from master: attached

Comment 1

[[e-mail] (:bhearsum)]

This is fixed now AFAIK