SS4 and NetBSD 10 stops

[Riccardo Mottola <riccardo.mottola%libero.it@localhost>]

Hi!

My SS4 has been compiling for aoc uple of days (not continuous uptimes,
I stop and resume) without issues, so I thought the stress test marked
it stable.
Remember that serial is not working and I don't keep Monitor&Keyboard
attached, since it is usually in a server pile.

Then I started building git-base which installs pcre2 dependency and I
find myself remotely disconnected.
I cannot reconnect via telnet.. so the system looks dead.

I Power cycle, clean build, restart... and again! and.... again!

Now I am attempting the same with keyboard&monitor attached.

However, I thought that maybe /var/log/messages could give a clue.

First restart (shows quite some uptime):

Jun 11 00:47:20 eowyn /netbsd: [  16.6991325] entropy: ready
Jun 11 00:47:34 eowyn savecore: no core dump
Jun 11 00:49:54 eowyn telnetd[1190]: ttloop:  read: Interrupted system call
Jun 11 14:00:08 eowyn syslogd[404]: restart
Jun 12 01:22:38 eowyn telnetd[3400]: ttloop:  read: Interrupted system call
Jun 12 01:23:44 eowyn telnetd[3878]: ttloop:  read: Interrupted system call
Jun 12 08:24:07 eowyn /netbsd: [ 113844.1002180] nfs server
ithil:/disk2/pkgsrc: not responding
Jun 12 08:33:15 eowyn /netbsd: [ 114391.8576060] nfs server
ithil:/disk2/pkgsrc: is alive again
Jun 12 14:00:02 eowyn syslogd[404]: restart
Jun 13 03:00:02 eowyn syslogd[404]: restart
Jun 13 14:00:03 eowyn syslogd[404]: restart
Jun 14 01:15:17 eowyn syslogd[554]: restart
Jun 14 01:15:17 eowyn /netbsd: [   1.0000000] Copyright (c) 1996, 1997,
1998, 1999, 2000, 2001, 2002, 2003,

second (did hang during the night...)

Jun 14 01:15:18 eowyn /netbsd: [ 128.2147475] entropy: ready
Jun 14 01:15:33 eowyn savecore: no core dump
Jun 14 08:25:28 eowyn syslogd[500]: restart
Jun 14 08:25:28 eowyn /netbsd: [   1.0000000] Copyright (c) 1996, 1997,
1998, 1999, 2000, 2001, 2002, 2003,


I suppose syslogd retarting from time to time would pose no issues?

The first time I can assume that the system had network issues since it
lost and regained connection to the nfs server (which is just 10 cm
below and the systems are on the same high-speed switch with short
patches...)
What are those ttloop messages?

Ideas? or is this just noise, maybe the system comes under heavy load.

Riccardo