My server has begun spontaneously checking parity this has happened randomly three times over the last three days. The first time It happened I noticed that I could not access the unRaid GUI through its IP address or any of the plugins on their ports. I figured that the server had crashed which would be very unusual as this machine has been very stable. The server was still running and the disk access lights were flashing.

 

A few minutes later web access started working again (without reboot). I accessed the main server page and noticed that a parity check had started. I let it complete and it did not discover any errors. Since then the parity check has kicked on at least once per day. I stopped it tonight and took a copy of the log (see attachment).

 

Please let me know if there is any other info that would help to figure this out. Thanks

 

Running unRaid Server Pro version 5.05

unRaid_log_10_Nov.txt

Nov 10 22:01:06 Tower sudo:    root : TTY=console ; PWD=/ ; USER=nobody ; COMMAND=/mnt/cache/apps/utserver/utserver -settingspath /mnt/cache/apps/utserver -configfile /mnt/cache/apps/utserver/utserver.conf -pidfile /var/run/utserver/utserver-8003.pid -logfile /mnt/cache/apps/utserver/utserver.log -daemon 2>&1 >/dev/null

Nov 10 22:01:11 Tower emhttp: unclean shutdown detected

 

5 seconds before it was loading utserver.  I would start looking there.

 

Hopefully one of the guru's will be along to assist!

Thanks for the analysis. I really don't understand why it would be detecting an unclean shutdown. At the time that that happened i wasn't doing any maintenance or work on the server.

 

 

 

 

Did you lose  power?  Do you have a lot of brownouts?  There was probably a surge in the power and it caused your server to shut down.  And your server BIOS is probably set up to reboot/restart on a power failure so you don't realize it has happened.

It sounds like a new undocumented feature

From the syslog it does not look like a spontaneous daily parity check.  It looks as though the server was booted.

Perhaps rebooted due to some power issue. Some of these messages only come up once during a boot up.

 

Nov 10 21:57:15 Tower syslogd 1.4.1: restart.
...
Nov 10 21:57:15 Tower kernel: NX (Execute Disable) protection: active
Nov 10 21:57:15 Tower kernel: SMBIOS 2.7 present.
Nov 10 21:57:15 Tower kernel: DMI: Gigabyte Technology Co., Ltd. To be filled by O.E.M./990FXA-UD5, BIOS FB 01/23/2013
...
Nov 10 21:57:52 Tower logger: Starting go script
Nov 10 21:57:52 Tower emhttp: unRAID System Management Utility version 5.0.5
Nov 10 21:57:52 Tower emhttp: Copyright (C) 2005-2013, Lime Technology, LLC
...
Nov 10 21:57:57 Tower emhttp: shcmd (24): /usr/local/sbin/emhttp_event array_started
Nov 10 21:57:57 Tower emhttp_event: array_started
Nov 10 21:57:57 Tower emhttp: Mounting disks...
...
Nov 10 22:00:13 Tower kernel: REISERFS (device sdh1): journal params: device sdh1, size 8192, journal first block 18, max trans len 1024, max batch 900, max commit age 30, max trans age 30
Nov 10 22:00:13 Tower kernel: REISERFS (device sdh1): checking transaction log (sdh1)
Nov 10 22:00:14 Tower kernel: REISERFS (device sdh1): replayed 610 transactions in 1 seconds
Nov 10 22:00:14 Tower kernel: REISERFS (device sdh1): Using r5 hash to sort names
...
Nov 10 22:01:11 Tower emhttp: unclean shutdown detected
Nov 10 22:01:11 Tower kernel: mdcmd (41): check CORRECT
Nov 10 22:01:11 Tower kernel: md: recovery thread woken up ...
Nov 10 22:01:11 Tower kernel: md: recovery thread checking parity...
Nov 10 22:01:11 Tower kernel: md: using 1536k window, over a total of 2930266532 blocks.

• The unclean shutdown message (and subsequent parity check) will happen anytime that the server is restarted without being cleanly shut down.  The proximity of that message to other log messages is likely to be irrelevant.
   
  You are loading and using a lot of plugins and it is always possible that one of those is crashing the system.  The way to check that would be to boot in Safe mode so that plugins are disabled and see if the problem still occurs
   
  Other possibilities are external factors such as
  • Loss of power
    
  • Power surges
    
  • Hardware issues such as failing power supply or motherboard
    

I agree that it looks like a restart but the instance from last night where the sys log was captured happened while I was watching plex and their was no interruption to the show that was being streamed from the server.

 

It just happened again about 10 minutes ago. The sys log again restarts as if the server has rebooted. I could not access the unraid web gui but I was able to telnet into the server and the user shares were accessible. This time my daughter who was watching a show on plex did say that her tablet told her that plex was suddenly no longer available.

 

I will disable auto restart under disk settings and see what happens next time this occurs.

syslog.txt

While 'proximity' of message may not matter, the time with proximity does. As I tried to point out earlier.

Plus some of these messages ONLY come up during a full reboot.

 

..
Nov 11 09:20:05 Tower syslogd 1.4.1: restart.
Nov 11 09:20:05 Tower kernel: Linux version 3.9.11p-unRAID (root@Develop) (gcc version 4.4.4 (GCC) ) #5 SMP Sat Dec 21 19:37:06 PST 2013
..
Nov 11 09:20:51 Tower kernel: REISERFS (device md4): replayed 285 transactions in 3 seconds
..

 

Your machine is crashing/rebooting/restarting due to some issue.

What that is, remains to be determined.

 

You may want to open a browser on a machine, click on the log button on the upper right and monitor the logs until it crashes/restarts again.

I would suggest you look at uptime via telnet. Make a note.

next time when you have question, do it again. you will see the server is rebooting.

 

Example:

 

root@unRAIDx:/# uptime

13:12:05 up 12 days, 13:35,  6 users,  load average: 0.00, 0.01, 0.05

OK thanks, seems pretty clear that it is crashing and rebooting then.

 

I am now monitoring the syslog via telnet using

 

tail -f /var/log/syslog

 

hopefully this will give some indication of what is causing it to crash.

 

 

After removing and reseating all cards and cables, checking drive health, running memtest for 2 days nothing seemed to jump out is being the cause of the random rebooting.

If anything the frequency of the reboots was increasing (sometimes 3 - 4 times per day).

 

Three days ago I decided to switch out the existing power supply (Corsair 550VX) for an older Corsair 500CX that I had in another machine. Since then there has not been another instance of random rebooting. Very happy about that. I was starting to think it might be the motherboard.

 

It really sucks because the 5 year warranty on that power supply just ran out in August. Anyone know it Corsair will show any leniency given how close I am to the warranty window?

 

Happy everything is working.