Hallo, sorry for the late update ... but here it is ...
On 10/29/2010 10:35 PM, Wietse Venema wrote: > If I don't see a credible report about warnings etc. in Postfix > logfiles, then that means that you are flying blind, and that needs > to be addressed first. Agreed. I did read the document about clocksource and timekeeping with VMware. I then configured each machine differently to see if there is any change in stability. > Regardless of why a process hangs, if it hangs then you should see > watchdog errors in the Postfix logs. If you don't see those then > either your virtual timer is busted, or your logging is busted. We guessed ourself that there was an issue with VMware or the config of the guest machine's kernel. I tried to reduce the number of CPUs on one machine to two, as well as setting the clocksource to "acpi_pm" on another one (left at eight cores); And the winner is: "less cpus". The machine with two virtual CPUs didn't crash once. Clocksource "acpi_pm" didn't help at all. So I now have all three instances running with two virtual CPUs for days without ANY issue whatsoever. As I don't need to performance of eight cores currently, I will leave it like that for now. Maybe any1 has more ideas based on the fact that the thing is stable with two cores now, but wasn't with eight. Regards, Christian
