So rolling out beta 15 has been a bit of a troublesome experience…
Our deadline farm is now much larger then previously, and auto updating only worked on about 30% of the whole farm.
On other machines one of these situations happened:
- slave was stalled, asking the launcher to restart the slave did not yield anything. Manually killing the launcher / restarting would trigger an update
- slave/launcher both running, launcher restarts the slave but does not update. Sometimes killing both would and restarting would trigger an update, but sometimes i actually had to reinstall deadline
- machine tries to restart the slave, but gets a database connection error, and doesnt do anything. After a while it is able to reconnect, and then updates itself.
- slave/launcher both running, but not responding to remote restart command. Manually shutting down the slave and starting again from the launcher would not update, until launcher is actually restarted
