Hello.
I'm running some small number of NVMM virtual machines. Sometimes one of
them just get killed with:
qemu: qemu_thread_create: Resource temporarily unavailable
dmesg shows:
pid 4381 (qemu-system-x86_): user write of 3221100864@0x76db1c5ce000 at
5314544 failed: 27
Number of process is low:
$ ps aux | wc -l
116
But number of LWPs is high:
$ ps aux -s | wc -l
1209
I'm running three instances of Firefox, they sum up 931 LWPs:
$ ps aux -s | grep firefox | wc -l
931
I tried to increase LWPs in sysctl with no success. It seems it is
hardcoded into the kernel code? Setting a new limit for threads (ulimit
-r) also doesn't work.
Well, not sure if number of LWPs is the problem, but I thought it was a
good idea to add this information here. It would explain why sometimes
one of my firefox instances gets killed randomly.
$ uname -a
NetBSD hope 9.2 NetBSD 9.2 (GENERIC) #0: Wed May 12 13:15:55 UTC 2021
mkre...@mkrepro.netbsd.org:/usr/src/sys/arch/amd64/compile/GENERIC amd64
Any help? Thanks!
