Michael Hannon wrote: > Greetings. We're running a mix of Scientific Linux (SL) versions on > Intel and AMD processors. All of our "production" systems are still > running SL 3.0.3. > > We noticed some time ago that the daily cron job to do yum updates > (/etc/cron.daily/yum.cron) was "hanging" routinely. I.e., there would > routinely be various yum-related tasks in the 'S' (sleeping) state on > our systems. These sleeping tasks would prevent the yum cron job from > running on succeeding days. > > We never found out the cause of the problem, but we did alleviate the > problem by removing yum.cron and scheduling our own daily yum updates. > Logically I wouldn't have expected there to be any difference in the > results produced by the two methods, but things did seem to work better > after removing yum.cron. > > Nevertheless, regardless of the yum voodoo that we try to apply, we > still are plagued by the "hanging-yum" problem. I was recently on > vacation for about a week, and I found half a dozen or so systems with > suspended yum-update jobs when I returned. > > I'd be glad to get suggestions for ways to eliminate this problem. Maybe > just updating to a newer version of SL would be enough to do the trick? > > Thanks. > > - Mike Hi Mike, The new yum-conf (in S.L. 3.0.5 and S.L. 4.1) has a fix for this. Basically it checks, and if yum has been running for 24 hours, the new yum.cron kills the old yum, and then procedes. We've been testing it, and this works quite well. Why does it hang? I've seen it mainly when we've had some type of network glitch or downtime. It seems like either yum or the server is waiting for the other to reply. I need to get this new yum out to errata (I consider it errata because it is a security issure if you arn't getting your updates.) hopefully by the end of the week, or beginning of next. Troy -- __________________________________________________ Troy Dawson [log in to unmask] (630)840-6468 Fermilab ComputingDivision/CSS CSI Group __________________________________________________