Skip site navigation (1)Skip section navigation (2)
Date:      Thu, 20 Sep 2012 13:05:17 -0600
From:      Gary Aitken <freebsd@dreamchaser.org>
To:        freebsd-questions@freebsd.org
Subject:   [solved] Re: 9.0 release not dead but barely breathing after idling
Message-ID:  <505B68ED.5070106@dreamchaser.org>
In-Reply-To: <503451D7.9040107@dreamchaser.org>
References:  <503451D7.9040107@dreamchaser.org>

next in thread | previous in thread | raw e-mail | index | archive | help
Finally.  Turns out I had done something really dumb;
the problem had nothing to do with screen blanking or X.

Because I send and receive mail via another server, 
I had disabled sending mail in rc.conf.
  don't do dat.
I wasn't getting the system messages targeted at root,
and eventually /var/spool/mqueue filled up.
With /var full, things don't work too well.
I didn't know /var was full because I wasn't getting the messages...
bad administrator.  bad! bad!  don't do dat ever again!

Gary

On 08/21/12 21:28, Gary Aitken wrote:
> Aargh...
> So my 9.0 RELEASE system no longer totally hangs when sitting idle...
> it seems to run quite a bit longer, waking up from screen blanking in general
> even after long (overnight) periods of sitting idle.  However, not always.
> 
> X (screen was allowed to blank after 10 min, I'm testing w/ that off now.)
> blanked the screen.
> I come back after a few hrs of the system doing nothing (leaving a lot of stuff open, esp in firefox) and the screen is blank (expected) but doesn't wake up.
> 
> I can ping from another machine, but not rlogin (no response).
> That seems weird.
> /var/log/messages shows no activity around attempted rlogin time
> Previously, before I turned off memory hole mapping in bios,
> it would go totally dead, but now it's clearly breathing.
> Power switch doesn't do a soft reboot,
> but I haven't tested it independently to see if it works at all.
> Will do that on next reboot.
> 
> Question:
> 
> If one does <ctl><alt>D to get out of X,
> how does that affect things?
> i.e. Since there is no active X display,
> what happens if a process tries to repaint?
> Does this effectively take display hardware out of the picture
> for troubleshooting?
> 
> Output from "last":
> 
> garya      pts/5    nightmare              Tue Aug 21 18:26 - 18:26  (00:00)
> garya      pts/3    :0                     Tue Aug 21 17:12   still logged in
> garya      pts/2    :0                     Tue Aug 21 17:12   still logged in
> garya      pts/0    :0                     Tue Aug 21 17:12   still logged in
> garya      pts/1    :0                     Tue Aug 21 17:12   still logged in
> garya      pts/4    :0                     Tue Aug 21 17:12   still logged in
> garya      ttyv0                           Tue Aug 21 17:08   still logged in
> boot time                                  Tue Aug 21 17:06
> garya      pts/3    :0                     Sun Aug 19 15:44 - crash (2+01:22)
> garya      pts/4    :0                     Sun Aug 19 15:44 - crash (2+01:22)
> garya      pts/2    :0                     Sun Aug 19 15:44 - crash (2+01:22)
> garya      pts/1    :0                     Sun Aug 19 15:44 - crash (2+01:22)
> garya      pts/0    :0                     Sun Aug 19 15:44 - crash (2+01:22)
> root       ttyv4                           Sun Aug 19 15:42 - crash (2+01:23)
> garya      pts/4    :0                     Sun Aug 19 15:15 - 15:41  (00:25)
> 
> I discovered the system was hung and rebooted around Aug 21 17:06.
> Why is no crash recorded on Aug 21?
>    The system was working (behaving normally) until at least ~ Aug 21 15:00
>    Since I could ping it around Aug 21 17:00,
>    but then did a forced power down in order to reboot,
>    shouldn't that show as a crash or something?
> Why is there no boot recorded soon after Aug 19 15:44?  (I did reboot)
> Why does the first entry for garya after boot show still logged in?
>    Is this because the records are based on the utx.log file,
>    and the system crashed, so it looks like I'm still logged in?
> 
> /var/log/cron shows:
> 
> Aug 21 13:11:00 breakaway /usr/sbin/cron[10699]: (operator) CMD (/usr/libexec/save-entropy)
> Aug 21 13:15:00 breakaway /usr/sbin/cron[10717]: (root) CMD (/usr/libexec/atrun)
> Aug 21 13:20:00 breakaway /usr/sbin/cron[10719]: (root) CMD (/usr/libexec/atrun)
> Aug 21 13:22:00 breakaway /usr/sbin/cron[10721]: (operator) CMD (/usr/libexec/save-entropy)
> Aug 21 13:25:00 breakaway /usr/sbin/cron[10733]: (root) CMD (/usr/libexec/atrun)
> Aug 21 17:10:00 breakaway /usr/sbin/cron[1878]: (root) CMD (/usr/libexec/atrun)
> Aug 21 17:11:00 breakaway /usr/sbin/cron[1882]: (operator) CMD (/usr/libexec/save-entropy)
> 
> So it appears the system went south around Aug 21 13:25
> 
> Anything else I should look at for hints?
> Any suggestions for how to narrow this down further other than:
>    disabling X screen blanking
>    <ctl><alt>D to get out of X prior to leaving machine idle
> 
> Thanks for any suggestions,
> 
> gary
> _______________________________________________
> freebsd-questions@freebsd.org mailing list
> http://lists.freebsd.org/mailman/listinfo/freebsd-questions
> To unsubscribe, send any mail to "freebsd-questions-unsubscribe@freebsd.org"
> 
> 




Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?505B68ED.5070106>