> yeh, I know you mean the cron service lock. But if they don't share the > service lock then they'll almost immediately hit the yum lock underneath > anyway and exit.
How could that work? Cron does not use the lock files at all, it can't serialize anything.. I'm sure hourly and daily yum-cron jobs have already competed for the yum.pid lock. OTOH, random sleep should help a lot there. > Or we could have the loop inside acquireLock() (the problem here is that > we'd have to be careful we don't stay around forever if someone leaves > a root "yum" around). Yes, there should be few retries, say 1 minute apart, to reduce the number of lock failures. _______________________________________________ Yum-devel mailing list Yum-devel@lists.baseurl.org http://lists.baseurl.org/mailman/listinfo/yum-devel
