can you get me a backtrace when the system panics? ive been trying to reproduce this locally without success.
cheers, dlg On 13/10/2010, at 11:00 PM, Marian Hettwer wrote: > Hi All, > > I'm wondering how I could configure scsi I/O timeout in OpenBSD. > I need to fiddle around with that since I'm using OpenBSD at work in some heavy (over)loaded ESX vmware cluster. From to time to the disk backend may response really slow. > A stock linux would remount the filesystem read/only, whereas my OpenBSD 4.7 boxes just paniced. (And Solaris 10 just kept on running. Probably some really high scsi i/o timeouts as default). > > In FreeBSD the sysctl kern.cam.da.default_timeout seems to do the trick. On Linux it's /sys/block/*/device/timeout. > > Is there an equivilant in OpenBSD? I couldn't find anything while searching the misc@ archives (using gmane.org). > > thanks in advance, > Marian > > PS.: please keep me CC'd, I'm not subscribed.
