On 28 Mar, Don Lewis wrote: > On 28 Mar, O. Hartmann wrote: > If I get a chance, I try booting my FreeBSD 11 machine with less RAM to > see if that is a trigger. I just tried cranking hw.physmen down to 8 GB on 11.0-CURRENT r297204, GENERIC kernel. /boot/loader.conf contains: geom_mirror_load="YES" kern.geom.label.disk_ident.enable="0" kern.geom.label.gptid.enable="0" zfs_load="YES" vboxdrv_load="YES" hw.physmem="8G" /etc/sysctl.conf contains: kern.ipc.shm_allow_removed=1 No /etc/src.conf and nothing of that should matter in /etc/make.conf. This is what I see after running poudriere ports -p whatever -u last pid: 2102; load averages: 0.24, 0.52, 0.36 up 0+00:06:54 14:13:51 52 processes: 1 running, 51 sleeping CPU: 0.0% user, 0.0% nice, 0.0% system, 0.0% interrupt, 100% idle Mem: 95M Active, 20M Inact, 1145M Wired, 39K Buf, 5580M Free ARC: 595M Total, 256M MFU, 248M MRU, 16K Anon, 14M Header, 78M Other Swap: 40G Total, 40G Free No swap used, inactive memory low, no interactivity problems. Next I'll try r297267, which is what I believe you are running. I scanned the commit logs between r297204 and r297267 and didn't see anything terribly suspicious looking. Are you doing any stateful firewalling? If the firewall connection state times out more quickly than the ssh keepalive interval, that would break connections.Received on Tue Mar 29 2016 - 19:26:37 UTC
This archive was generated by hypermail 2.4.0 : Wed May 19 2021 - 11:41:03 UTC