My build machine ("freebeast") spends most of the time powered off.

One of my "always on" machines has a crontab entry for 23:47 to use
/usr/local/bin/wol (from ports/net/wol) to wake it up in time to do some
periodinc "daily" things, update its local mirror of the SVN repos, and
update it ports working copy; after I've updated stable/10 & head on it,
I set it to boot from the stable/10 slice & power it off -- and the
cycle repeats.

At least, that's what happened with this machine's predecessor for
several years, and with this one since its deployment several months
ago, until yesterday morning.

Yesterday morning, the machine did not power on; I figured someone had
managed to manually power it off (vs. "shutdown -p"), took evasive
action, and resumed normal operations.

Upon a recurrence this morning -- and after the evasive action &
resumption of normal operations, I decided to test a bit.  The machine
was last running:

FreeBSD 11.0-CURRENT #1897  r290667M/290668:1100089: Wed Nov 11 04:45:58 PST 
2015  amd64

I manually ran the /sur/local/bin/wol command... and Nothing Happened.

Went over to the machine; no lights.  Manualy turned it on (booting

FreeBSD 10.2-STABLE #1851  r290668M/290668:1002501: Wed Nov 11 04:20:05 PST 
2015  amd64

-- just to be clear); once it came up, "shutdown -p now".  After seeing

acpi0: Powering system off

on serial console, re-issued /usr/local/bin/wol; machine came right up.

OK.  So /usr/local/bin/wol itself seems to be working, and freebeast's
hardware also seems OK.

Checking the log of builds for head on that machine, the last several
entries are:

FreeBSD 11.0-CURRENT #1892  r290439M/290439:1100086: Fri Nov  6 05:12:44 PST 
2015  amd64
FreeBSD 11.0-CURRENT #1893  r290486M/290488:1100087: Sat Nov  7 07:48:26 PST 
2015  amd64
FreeBSD 11.0-CURRENT #1894  r290542M/290542:1100089: Sun Nov  8 06:19:02 PST 
2015  amd64
FreeBSD 11.0-CURRENT #1895  r290606M/290609:1100089: Mon Nov  9 04:44:14 PST 
2015  amd64
FreeBSD 11.0-CURRENT #1896  r290647M/290647:1100089: Tue Nov 10 04:37:17 PST 
2015  amd64
FreeBSD 11.0-CURRENT #1897  r290667M/290668:1100089: Wed Nov 11 04:45:58 PST 
2015  amd64

and since I'm pretty sure I would have recalled had the "wake on
LAN" failed on Monday (and I don't recall that), it seems likely that
some change made that day (thus, a commit between r290542 - r290606)
is what would have done it.

But a quick perusal of
<> doesn't show
anything especially like a "smoking gun" -- to me, anyway.

Can anyone else confirm or refute my observations?  Or suggest a
hint?  I'll try narrowing it down myself, but I need to do it during
times I'm at home (so I can manually power the machine back up when
it fails to respond to WoL), so it may be a few days before I can
accomplish much that way.

David H. Wolfskill                    
Those who would murder in the name of God or prophet are blasphemous cowards.

See for my public key.

Attachment: signature.asc
Description: PGP signature

Reply via email to