Announcement

**Snowhog** · Jul 28, 2010, 05:31 PM

Re: Sponteneous Reboots

Spontaneous reboots are almost always a result of a power loss or a heat problem (CPU). How hot is your PC running?

**Don** · Jul 29, 2010, 06:16 AM

Re: Sponteneous Reboots

Thanks for the response, Snowhog. I think you are probably correct. We haven't had any power problems at work, so heat is a likely culprit. I'm back at the machine, but I don't know how to check its temperature. Sensors-detect says there are no sensors. Can you point me in the right direction? (Use small words; I'm still very new to Linux.)

**dibl** · Jul 29, 2010, 07:31 AM

Re: Sponteneous Reboots

Don, did you first install the package "lm-sensors"? If yes, and

Code:

sudo sensors-detect

reports no sensors, then that means your computer does not have any embedded thermal sensors for which lm-sensors has a driver.

In that case, the best bet is to carefully open the case, in a location where you can use compressed air and/or a vacuum (with plastic, not metal nozzle) to remove dust accumulations on fans and/or filters. Be careful not to touch the electronics with anything that is conductive, and be cautious about how fast the fan might get spun with compressed air -- don't overdo it.

**oshunluvr** · Jul 29, 2010, 07:45 AM

Re: Sponteneous Reboots

To add a bit to Snowhog's response in case tohers are viewing this thread... it may also be a perceived heat problem.

What I mean is linux is trying to protect your machine because it thinks it's too hot - it may not be actually overheating. This feature is a part of ACPI and can be in error.

In my case, I could get my machine to run for about 3 minutes until it would shut off due to over heating. But my bios temp read out was fine. Turning ACPI allowed my machine to stay on, but I lost other features (like hyper-threading).

In my case the final solution was a BIOS update that corrected the temp readout. Others I've read about were able to adjust the temp settings to allow their system to run "hotter".

You might try booting with ACPI off and seeing if makes it behave differently. Obviously - verify first that you are not actually over heating. Also, in my case - every shutdown due to this was logged. The line looked something like:

May 15 00:23:45 media kernel: [ 3391.624000] ACPI: Critical trip point
May 15 00:23:45 media kernel: [ 3391.624000] Critical temperature reached (71 C), shutting down.

I don't see anything in your log that shows that - leading me to believe the issue is somewhere else or is a temp problem not detected by ACPI.

To your issue Don: How long has it been since your done any basic maintenance to your machine? You might try looking to see if dust has built up in your fans. Also, opening up your case and running the machine with it open may reveal a fan that has quit and will also allow your system to run cooler. If it doesn't reboot with it open you may have discovered your problem. Also - while it's open and before you turn it on try re-seating all your cards and memory sticks. Sometimes heat expansion will allow cards or memory to wiggle out of their slots a bit.

Last thought on this topic: Your power supply may be failing. The power loss might be internal rather than external. I don't know how you would check this without just changing it.

**vinnywright** · Jul 29, 2010, 07:58 AM

Re: Sponteneous Reboots

Originally posted by dibl

Don, did you first install the package "lm-sensors"? If yes, and

Code:

sudo sensors-detect

reports no sensors, then that means your computer does not have any embedded thermal sensors for which lm-sensors has a driver.

In that case, the best bet is to carefully open the case, in a location where you can use compressed air and/or a vacuum (with plastic, not metal nozzle) to remove dust accumulations on fans and/or filters. Be careful not to touch the electronics with anything that is conductive, and be cautious about how fast the fan might get spun with compressed air -- don't overdo it.

yes the last time some one brought me a box that was mistereusley shuting down I opend the case and saw a maze of spider web's and what looked like a big gray cotton ball in the CPU fan (I dought it was even spining) the fan had to be removed frome the heet sink to alow cleening of the fins .

but once it was all cleend up both the dirt and virus's it's ben runing just fine ........well exept that I had to leave windows on it .........LOL

VINNY

**oshunluvr** · Jul 29, 2010, 08:29 AM

Re: Sponteneous Reboots

Co-worker of mine is part-time Geek Squad - says he seen systems that looked like they would catch fire if turned on!

**Don** · Jul 29, 2010, 11:33 AM

Re: Sponteneous Reboots

I regret to report that the inside of the case was remarkably (some might say supernaturally) dust free (I vacuumed it out anyway), and the fan is working properly. I will check out BIOS updates. As for the power supply, I have another one on hand that I can sub in; if the problem continues after the update, that's what I'll do. Meanwhile, I worked on the machine for 3.5 hours this morning in OOo Writer and had no incidents. It's these sporadic things that drive all of us nuts.

**dibl** · Jul 29, 2010, 11:40 AM

Re: Sponteneous Reboots

I once had the task of troubleshooting a PC that sporadically rebooted -- the symptoms were about the same as this case. Eventually (and after a painful amount of expense and degradation of customer satisfaction), I discovered a cold solder joint in the PSU. It was closed when cold, so would boot and run normally for awhile. When it warmed up, and depending on the power load, it would open and trigger the reboot. So there's another angle to consider.

**wizard10000** · Jul 29, 2010, 12:18 PM

Re: Sponteneous Reboots

Originally posted by dibl

I once had the task of troubleshooting a PC that sporadically rebooted -- the symptoms were about the same as this case. Eventually (and after a painful amount of expense and degradation of customer satisfaction), I discovered a cold solder joint in the PSU. It was closed when cold, so would boot and run normally for awhile. When it warmed up, and depending on the power load, it would open and trigger the reboot. So there's another angle to consider.

Yup. It could also be RAM, motherboard, all kindsa stuff.

A Intel-architecture CPU depends on a "power_good " signal to the processor to stay booted. Without that signal the timer chip will constantly reset the CPU until it receives that power_good signal. Older (non-ACPI) systems used to use the reset button to deassert the power_good signal to cause the processor to reset.

**mhumm2** · Aug 01, 2010, 07:33 PM

Re: Sponteneous Reboots

Does ACPI have any type of log file to analyze or is syslog the only one?

**Don** · Aug 03, 2010, 03:49 PM

Re: Sponteneous Reboots

I'm beginning to think that the problem is really with OOo Writer rather than being more systemic. The reboots happen only when I'm using that program, and there's plenty that I do outside that program that I should have experienced such a reboot at least once. Even when I'm using the vm, I've never had this happen.
So I'm wondering whether I need to take OOo out and then reinstall it. Does that make any sense?

**Snowhog** · Aug 03, 2010, 06:51 PM

Re: Sponteneous Reboots

Originally posted by Don

So I'm wondering whether I need to take OOo out and then reinstall it. Does that make any sense?

Stranger things have/do happen. If you are confident that OOo is the cause, removing it (sudo apt-get purge package_name) might be a valid solution. It won't hurt to try - your data files are in your users home directory.

**oshunluvr** · Aug 04, 2010, 05:29 AM

Re: Sponteneous Reboots

Have you looked in your logs at all? IMO unless it's directly hardware related - there will be something in the log files if a reboot is software related...

**Don** · Aug 04, 2010, 09:56 AM

Re: Sponteneous Reboots

I dearly hope this is the log file that oshunluvr wanted to see (because otherwise I don't know where to find it). Note that at 11:26:14, there was a spontaneous reboot. (I'm not saying that because I can read the log intelligently; that's just when I had to login again.)

2010-08-04 08:17:01 745 CRON[13521] (root) CMD ( cd / && run-parts --report /etc/cron.hourly)
2010-08-04 09:17:01 745 CRON[13525] (root) CMD ( cd / && run-parts --report /etc/cron.hourly)
2010-08-04 10:17:01 745 CRON[13703] (root) CMD ( cd / && run-parts --report /etc/cron.hourly)
2010-08-04 11:17:01 745 CRON[13788] (root) CMD ( cd / && run-parts --report /etc/cron.hourly)
2010-08-04 11:26:14 745 kdm[743] X server for display :0 terminated unexpectedly
2010-08-04 11:26:14 745 acpid client 11619[0:0] has disconnected
2010-08-04 11:26:14 745 acpid client connected from 13812[0:0]
2010-08-04 11:26:14 745 acpid 1 client rule loaded
2010-08-04 11:26:15 745 kdm_greet[13814] Cannot load /usr/share/kde4/apps/kdm/faces/.default.face: No such file or directory
2010-08-04 11:26:26 745 anacron[13965] Anacron 2.3 started on 2010-08-04
2010-08-04 11:26:26 745 anacron[13965] Normal exit (0 jobs run)
2010-08-04 11:26:26 745 kernel [510668.593648] CPU0 attaching NULL sched-domain.
2010-08-04 11:26:26 745 kernel [510668.593656] CPU1 attaching NULL sched-domain.
2010-08-04 11:26:26 745 kernel [510668.616143] CPU0 attaching sched-domain:
2010-08-04 11:26:26 745 kernel [510668.616149] domain 0: span 0-1 level CPU
2010-08-04 11:26:26 745 kernel [510668.616154] groups: 0 1
2010-08-04 11:26:26 745 kernel [510668.616163] CPU1 attaching sched-domain:
2010-08-04 11:26:26 745 kernel [510668.616167] domain 0: span 0-1 level CPU
2010-08-04 11:26:26 745 kernel [510668.616171] groups: 1 0

If there is something else I need to check, please let me know what it is and where to find it.

It still is the case that this only happens when I'm using OOo Writer. (Murphy's Law strikes again: I'm racing to finish an article.)
Thanks for the help.

Announcement

[SOLVED] OpenOffice crashes X Server

[SOLVED] OpenOffice crashes X Server

Comment

Comment

Comment

Comment

Comment

Comment

Comment

Comment

Comment

Comment

Comment

Comment

Comment

Comment