Can someone help me....my server keeps restarting by itself and I don't know why. For the last week and a half I noticed it was locked up once or twice and had to power it down via the power button. When it came back up it did a parity check as was expected. Now it seems to just be restarting by itself and the way I know this is by the last day a parity check was run. I noticed Sunday that a parity check was run that day. Today I look and see one was run yesterday morning.

Now we are doing construction in the house and I have been cleaning drywall dust off the front grill everyday. I thought at first it was heat related but I check the SMART info under 190 Airflow_Temperature_Cel and the MAX. was 23 on most of the drives...none showed a dangerous level.

 

I attached my syslog and if there is anything else I can provide please let me know.

 

My server has been running rock solid for over a year now....until now

 

Thanks for any help!!!

 

 

Edit: Sorry I am running 4.3.3 and parity check shows no errors.

system_log.txt

What kind of PSU do you have?  Have you added any new drives or components lately?  Your symptoms are common for a dying or overloaded PSU.  Perhaps your PSU fans are clogged with dust from the construction, and the whole thing is overheating.

I didn't see anything really stand out here.

Turn off cache_dirs for a few days and see how it goes.

 

You need to check the syslog every few hours for the next couple of days to see if anything crops up.

 

Another choice is rsyncing it to the flash a couple of times a day.

 

I could understand it locking up, failing to do any network I/O, but the only thing that would make it reboot is  a power dip (you have a UPS right) or an issue with the power supply.

 

Another thought - perhaps using heavy power tools through the house mains is causing 'brown outs' throughout the rest of the house.  A UPS would eliminate this possibility.

At times like that, the first suspect (and the one that's most easily testeded) is RAM.

From the boot menu select "Memtest86+".  Let it run for a few hours.  Preferrably, leave it running overnight.

 

I attached my syslog...

That syslog is a brand new one from AFTER the server restarted. It doesn't give any clues as to what happened just before the restart.

 

In unRAID the syslog is kept in RAM. If the server crashes that syslog is gone!

So, apply the following patch, and it will mirror your syslog to the USB flash disk.

mv -f /boot/syslog.current /boot/syslog.last 2>/dev/null
cat /var/log/syslog >/boot/syslog.current
echo '*.*  /boot/syslog.current' >>/etc/syslog.conf
echo '*.*  /dev/tty12' >>/etc/syslog.conf
/etc/rc.d/rc.syslog restart
setterm -term linux -powersave off -powerdown 0 -blank 0 </dev/tty12 >/dev/tty12

Open a telnet session to your server, and Copy/Paste the above lines into the telnet window.

Also, add the above six lines to your 'go' script, so they can be executed every time the server boots up.  

 

Now when the server crashes, you'll always have the last syslog on the flash key, showing all critical events from just before the crash.  You can  plug in the flash disk into any other computer and examine the syslogs on it.  The interesting file on the flash disk will be called "syslog.current".  But if you've rebooted the unRAID server, then that same log (from before the reboot) will be named  "syslog.last".

http://lime-technology.com/forum/index.php?topic=6950.msg67511#msg67511

 

Note: remember to always "safely eject" the flash disk after you've examined it on another computer.

 

The above patch also mirrors your syslog to virtual terminal tty12 on the console.  If you have a monitor and keyboard attached to your server, press Alt-F12 to switch to tty12.  When the server freezes, you may still have the latest syslog entries visible on the screen.

 

Note: After you find out what the problem is and resolve it, you may want to remove that patch from the 'go' script. No need to wear out the flash disk unnecessary.

 

 

hi

i am having the same problem.

have you managed to pinpoint the reason for your restarts?

The only time I have had spontaneous restarts was when capacitors on the motherboard had gone bad.  That creates weird and transient voltage problems.

i am having the same problem.

have you managed to pinpoint the reason for your restarts?

It will be more accurate to say: you're having the same symptoms. Your actual problem may be different than his.

Any one of a wide variety of problems can be the reason for your headaches.  Start with the troubleshooting section of the wiki:

http://lime-technology.com/wiki/index.php?title=Troubleshooting

 

thanx, i am gonna have a look..

it really is very frustrating because there is nothing in the syslog that can help and it keeps restarting (sometimes every few hours, sometimes after 3 or 4 days) for no apparent reason. and everytime it starts doing a parity check because of the abrupt shutdown...

i hope i will find a solution in the troubleshooting section...

thanx, i am gonna have a look..

it really is very frustrating because there is nothing in the syslog that can help and it keeps restarting (sometimes every few hours, sometimes after 3 or 4 days) for no apparent reason. and everytime it starts doing a parity check because of the abrupt shutdown...

i hope i will find a solution in the troubleshooting section...

Are you on a UPS?

it really is very frustrating because there is nothing in the syslog

Are you looking at the syslog that's from before the reboot? 

Because that one is not the same file as your current syslog.  (see the post above)

 

@WeeboTech:

no i am not on ups

 

@purko:

i have looked at previous syslogs. there is no indication as to why these restarts occur. obviously the pc restarts suddenly and no info is written on the syslog (for example, the pc may suddenly restart at 20.00 and there is nothing in the log after 17.00)

additionaly, there is nothing that would seem not normal in the whole log

 

PS: i have run a memtest and checked the cpu's temp (although i dont know how accurate the sensor is) but everything seems fine. i guess it should be narrowed down to the PSU or the mobo.

unless it has something to do with the upgrade to version 4.5.4, but that seems impropable since the problems started appearing much later....

@WeeboTech:

no i am not on ups

 

@purko:

i have looked at previous syslogs. there is no indication as to why these restarts occur. obviously the pc restarts suddenly and no info is written on the syslog (for example, the pc may suddenly restart at 20.00 and there is nothing in the log after 17.00)

additionaly, there is nothing that would seem not normal in the whole log

 

PS: i have run a memtest and checked the cpu's temp (although i dont know how accurate the sensor is) but everything seems fine. i guess it should be narrowed down to the PSU or the mobo.

unless it has something to do with the upgrade to version 4.5.4, but that seems impropable since the problems started appearing much later....

 

Your server should be on a  UPS.

How do you know the problem is not external power?

 

@WeeboTech:

no i am not on ups

 

@purko:

i have looked at previous syslogs. there is no indication as to why these restarts occur. obviously the pc restarts suddenly and no info is written on the syslog (for example, the pc may suddenly restart at 20.00 and there is nothing in the log after 17.00)

additionaly, there is nothing that would seem not normal in the whole log

 

PS: i have run a memtest and checked the cpu's temp (although i dont know how accurate the sensor is) but everything seems fine. i guess it should be narrowed down to the PSU or the mobo.

unless it has something to do with the upgrade to version 4.5.4, but that seems impropable since the problems started appearing much later....

 

Your server should be on a  UPS.

How do you know the problem is not external power?

 

 

well, i have another pc on the same outlet and it never has any problems.

additionally, i use these surge/power spike protection plugs...

 

 

well, i have another pc on the same outlet and it never has any problems.

additionally, i use these surge/power spike protection plugs...

 

That does not indicate that a power fluctuation is not the problem.

Surge and power spike protectors only protect in one direction (voltage spike).

a UPS protects in multiple directions, Spike, Dip, Fluctuations of any kind.

 

Age of board (capacitors), Heat, Power all come into play here.

 

well, i have another pc on the same outlet and it never has any problems.

additionally, i use these surge/power spike protection plugs...

 

I'm not sure I'd consider your other PC as a fail safe test that "brown outs" aren't a problem.  Your unRAID server, particularly if PS is close to edge, could simply be more sensitive to power fluctuations than your PC.  I agree with the others, put it on a UPS.

ok, maybe you're right

i'm gonna order a ups (i was planning to anyways) and get back to you...

The average ups will compensate for below average psu.  Seriously, I have been using ups power for all my computers since before y2k.  I still have a couple machines that date back to 2001 running 24x7.  A ups will add years to all your computer equipment.

Wow,

I just read this and honestly I wouldn't even load up my data until I picked up a UPS and had it powered up and plugged in. I figure a $99 UPS to protect 100's if not several 100's of hours ripping DVD's and other things.

hi everybody

i had ordered a ups about 10 days ago but the shop was stalling without any explanation so i cancelled it. so i am looking again for a ups. i understand that apc is best supported so i am thinking of one of the following 2:

http://www.amazon.com/Back-ups-RS-330-Watts-550VA/dp/B002RL0CKI/ref=sr_1_2?ie=UTF8&s=electronics&qid=1279813152&sr=8-2

or

http://www.amazon.com/APC-BR800I-800VA-230V-UPS/dp/B00013MRS4/ref=sr_1_1?s=pc&ie=UTF8&qid=1279813389&sr=1-1

Where i live, the cheapest price i can get is 110 euros for the first one, and 160 euros for the 2nd one

 

I will connect my main PC and the unraid server on the ups.

Is it worth spending the extra 50 euros or should i go with the cheap one?

I am not sure where are you located in the euro zone but you can try to buy the APC from your local DELL site. (www.dell.fr, dell.de, dell.at etc....)

Dell does sell the APCs and often they have promotions with unbeatable prices.

Also Dell may ship for free... (at least it does here in Canada) and righ now they have one 550VA 330W APC unit for the equivalent of like 35-40 euros.

 

 

I am not sure where are you located in the euro zone but you can try to buy the APC from your local DELL site. (www.dell.fr, dell.de, dell.at etc....)

Dell does sell the APCs and often they have promotions with unbeatable prices.

Also Dell may ship for free... (at least it does here in Canada) and righ now they have one 550VA 330W APC unit for the equivalent of like 35-40 euros.

 

 

 

i live in greece and these are the best prices i can get. Most big sites wont ship electronics such as ups outside the country they are located (or, at least, to Greece)

hi everybody

i had ordered a ups about 10 days ago but the shop was stalling without any explanation so i cancelled it. so i am looking again for a ups. i understand that apc is best supported so i am thinking of one of the following 2:

http://www.amazon.com/Back-ups-RS-330-Watts-550VA/dp/B002RL0CKI/ref=sr_1_2?ie=UTF8&s=electronics&qid=1279813152&sr=8-2

or

http://www.amazon.com/APC-BR800I-800VA-230V-UPS/dp/B00013MRS4/ref=sr_1_1?s=pc&ie=UTF8&qid=1279813389&sr=1-1

Where i live, the cheapest price i can get is 110 euros for the first one, and 160 euros for the 2nd one

 

I will connect my main PC and the unraid server on the ups.

Is it worth spending the extra 50 euros or should i go with the cheap one?

 

Since you don't already use a UPS I'm assuming you don't have long blackouts in your area.  So the battery doesn't have to provide a long run-on-battery time for you.  Your main reason is to eliminate the nasty brown outs that occur all the time even in good power areas.  Those are what usually kill power supplies.  And any UPS will smooth out the spikes because of the filtering.

 

That gets it down to how many volt-amps you will need.  A 6 drive server versus a 18 drive server.  I would say the small UPS should handle 8-10 drive system.  The UPS will have battery outlets and surge-protector outlets.  Put the server on battery and other items like the monitor in surge-protection.

 

 

It took long for me to get back to you because i was away for my vacation.

 

I also wanted to try my new UPS for some time.

 

It turns out that most of you people, who suggested that the reason my computer is restarting might be the unstable power supply in my area of residence were right. After plugging my UPS i have not had any more restarts

 

So i am just posting this in order to help anyone else who might have the same problem.

btw, i must mention that my PSU is quite new and of good quality, still it did not prevent the restarts...

 

I have some questions as to how the unraid server can see my APC UPS (cause for the time beiing, it cannot) but i will get back to you when i have some more time...

 

Thanx everyone for the help and advice...