[mythtv-users] Backend powered off/shutdown unexpectedly!

[Ian Evans]

On Thu, Jan 30, 2014 at 3:25 AM, Mike Hodson <mystica at gmail.com> wrote:

> I should also mention if there are overheat events, sometimes those show
> up as kernel machine check exceptions or as thermal throttle messages.
>
> They should be logged as well, and I also forgot to mention these logs can
> be read in realtime with the 'dmesg' command.
> On Jan 30, 2014 1:20 AM, "Mike Hodson" <mystica at gmail.com> wrote:
>
>> Hello Ian,
>>
>> There's actually a few things that could cause a power off..
>>
>> First worth mentioning, as you say you recently started having issues
>> after adding more hardware and this more power drain on certain parts of
>> the power supply, that there may be sizing issues or perhaps a power supply
>> going bad depending on its age and usage margin between rated output and
>> actual usage ... Or you may be one of the unlucky ones with capacitors that
>> pop open and cause the pay to die an early death.
>>
>> Other parts would be an ACPI power off from the power button if something
>> bumped it... This should be logged if kernel logs are enabled. I don't know
>> which release of Ubuntu you are on but newer ones may log different than
>> older ones. The version would be helpful, but often these are in /var/log
>> and usually called messages or a kernel log file name. I'm not an Ubuntu
>> person myself so perhaps someone more versed could chime in on this
>> aspect... I use syslog-ng on funtoo myself...
>>
>> Finally there is always overheating potential, which cab be logged via a
>> sensors program, usually included or easily installed. I know of
>> LM_sensors, which can be run with 'sensors' from a command line or ssh once
>> the appropriate settings are found. There is an auto configure script if I
>> remember correctly, its been a while.
>>
>> Hope this helps!
>>
>> Mike
>> On Jan 29, 2014 7:10 PM, "Ian Evans" <dheianevans at gmail.com> wrote:
>>
>>> Combo MythTV/Ubuntu/hardware question here.
>>>
>>> Just got in to discover that our backend was off for the last three
>>> hours. 2nd time that has happened since Xmas when we added a new drive.
>>> Forgot to write then because I was working on a deadline. However, the
>>> system has been rock solid otherwise, lots or recordings, no video
>>> glitches, etc.
>>>
>>> Won't be able to reboot and check BIOS until after midnight (recordings)
>>> but can SSH in and check logs. So some questions:
>>>
>>> 1) BIOS is/was set to boot on power. Since this has only happened in
>>> two+ yrs twice since we added a drive at Xmas, is there a way that setting
>>> could have been bumped or switched off on its own? Didn't make any changes
>>> to BIOS when I added the drive but just covering my bases.
>>>
>>> 2) If it's something like "shutting down for safety" (detected
>>> overheating/etc) is there any log, Myth or Ubuntu, that would show some
>>> message that translates to "Emergency! Powering down!" so I can see what
>>> happened?
>>>
>>> 3) I don't have Wake-on-LAN set, but is it just wake, or can it power on
>>> the system as well? I run a website so perhaps I could have that server
>>> ping our dyndns every five minutes or so and send a wake up call if the
>>> system is down,
>>>
>>> Thanks for any help. WAF is at risk! :-)
>>>
>>>
Seems to be a heat event. Rebooted to get to the BIOS (Award 6.00pg) and
can't seem to find anything about WOL there. Suddenly the box shut down.
Rebooted. Went back to the bios and the CPU was running at 113C. The CPU
fan was showing 0 RPM so either it died or the connector got loose perhaps
when I added the other drive.

Now to get it off the entertainment unit...which is a task and a half.
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://www.mythtv.org/pipermail/mythtv-users/attachments/20140130/352535ba/attachment-0001.html>