On Sun, Sep 09, 2007 at 22:59:35 +0200, Michael Schwendt <mschwendt@xxxxxxxxx> wrote: > On 09/09/2007, Bruno Wolff III wrote: > > > > There do appear to be some kernel problems relating to I/O deadlocks that > > can be triggered by yum. However in my case, reboots have always temporarily > > fixed the problem. > > Are such kernel bugs known in bugzilla already? > Does it affect only recent kernels in F7? Is it fixed in F8 development? I am seeing what appear to be I/O deadlocks in in F7. The symptons are a bit different than I saw in FC5. (In FC5 the time to lock up was relatively consistant for a particular machine, in F7 it seems more tied to heavy disk/memory use.) I have a bug open about the issue for FC5 and have added comments to some other bugs in Redhat's bugzilla. I don't have the problem on my home machine, but that machine has a lot more memory (the problem machines only have 512MB, my home machine as 2GB) and a lot less disk activity (its not an email or web server). I haven't tried F8 yet, as I don't have a convenient place to play with it right now.
