On Sun, Oct 24, 2021 at 11:26:40AM +0100, Stuart Henderson wrote: > On 2021/10/24 10:01, Jan Stary wrote: > > On Oct 23 19:00:13, [email protected] wrote: > > > Can you go into any more detail on "process will get killed" or "crashed"? > > > Core dumps, backtraces, messages on console/logs/dmesg? > > > > Console just says 'Killed', as in > > > > $ pstree > > Killed > > > > There is no core dump, nothing in dmesg, nothing in logs > > that I could find. > > > > But: > > > > > Does the swap partition sd0b actually exist, > > > or are you by any chance running these machines swapless? > > > > Indeed, I am running without swap. > > Ah yes AFAIK that is not really supported.
However, the specific problem Jan describes is not reproducable on amd64. We've been running almost all of our amd64 OpenBSD machines swapless for over 15 years, with no problems whatsoever. So it's either specific to arm64, or it's a common issue in the uvm code which for some reason is never triggered on amd64, (or i386 as far as I am aware).
