After leaving all my data migrating to my shiny new unraid install last night, I woke up to "Array stopped: unclean shutdown detected".

 

At first I thought it must have been a power cut but the other servers were fine, in fact on the one that was copying the files I can see that the file transfer seemed to have been interrupted at bang on midnight.

 

I've downloaded the diagnostic logs to see if they could shed any light on anything but I can't see anything that would indicate a shutdown or startup procedure - I'm stumped! Any help greatly appreciated!

The evidence you have given us

Sep 11 16:01:22 ODIN rsyslogd: [origin software="rsyslogd" swVersion="8.6.0" x-pid="1191" x-info="http://www.rsyslog.com"] start
...
Sep 11 20:14:47 ODIN emhttp: import flash device: sda
Sep 12 01:31:22 ODIN emhttp: shcmd (28): rmmod md-mod |& logger
...
Sep 12 02:29:41 ODIN emhttp: import flash device: sda

indicates the server was up at midnight Sep12.

 

Seems like the evidence or your description must be incorrect or incomplete in some way.

Thanks for your reply trurl, but that just seems to confuse things further - does that essentially mean that the server never lost power or shutdown, but still somehow managed to get into an unclean shutdown state?

 

Is it perhaps that I'm misunderstanding that unclean shutdown meant the computer was physically restarted, whereas it may have meant that the array itself was shutdown due to an error?

Beyond your description, we don't have any evidence of an unclean shutdown.

 

Is your server time set correctly? Maybe the syslog is not showing us your local time.

 

If you look at the upper right of the webUI it should tell you the Uptime.

Time is set correctly, and it was actually the system uptime which first led me to believe it was a shutdown - as the uptime is 14 hours and 47 mins, and the time is 14:48 (see attached).

 

EDIT: Timezone was set incorrectly (my mistake), was set to Pacific not UK.

So where did you see

"Array stopped: unclean shutdown detected"

In the red area attached, written in orange. When I started the array it immediately started a parity check which is still ongoing.

 

And there is no parity check in your syslog either.

After leaving all my data migrating to my shiny new unraid install last night, I woke up to "Array stopped: unclean shutdown detected".

Are you migrating from one unRAID to another? If so, are you sure you don't have them mixed up as far as which one had the unclean shutdown?

No from Windows server.

 

But as mentioned above (my bad) the timezone was set incorrectly, meaning the startup at 16:01 would have actually been 00:01.  Assuming you can see no reason in the logs why this occurred, I'm guessing this must of been a hardware/power issue?

 

Edit: fresh logs attached

No from Windows server.

 

But as mentioned above (my bad) the timezone was set incorrectly, meaning the startup at 16:01 would have actually been 00:01.  Assuming you can see no reason in the logs why this occurred, I'm guessing this must of been a hardware/power issue?

 

Edit: fresh logs attached

Yes. Since your syslog starts at midnight (after the adjustment) then it was a hardware/power issue.

 

You should have your server on UPS. Is your BIOS set to restart on power restore? Not sure that is a good setting for your server.

 

Could have also shutdown due to some hardware issue like CPU overheat but power seems most likely.

It was the timing (exactly midnight) and the fact nothing else lost power that was throwing me, but it is the most obvious culprit.

 

You're right in that it should be on a UPS, and it will be eventually when I start storing important stuff on there - and when money permits...

 

Thanks for all your help on this wild goose chase!

It was the timing (exactly midnight) and the fact nothing else lost power that was throwing me, but it is the most obvious culprit.

 

You're right in that it should be on a UPS, and it will be eventually when I start storing important stuff on there - and when money permits...

 

Thanks for all your help on this wild goose chase!

The timing does seem a little suspicious. Sure you don't have something that triggered that?

Ok the plot thickens - Happened again, came back to the same message "Array stopped unclean shutdown detected". Chalked it up to bad luck, started looking at UPS units.

 

Then while sat next to it last night (while it was doing a parity check) I noticed I couldn't get on the web interface, SSH was not responding, and the unit had no output to a monitor I connected. As it was late last night and I couldn't get a response for nearly ten minutes, I hard powered it down and left it over night.

 

Today I removed all the drives (I'm getting concerned about data integrity now) and ran memtest for over an hour with no faults. My next thought was to power up the unit and get the log files off with the drives still removed and then turn it off again while I go through them. However the machine thew a kernel panic as soon as it started to boot and hung. (See attachment). While checking the BIOS i also note than auto restart on power failure is actually off, meaning the other instances where this happened is even more peculiar as the machine decidedly did not lose power, or else the machine would have still be powered off when I noticed.

 

I'm at a bit of a loss now, I noticed the error message mentions FAT, could there be a problem with the USB stick?

Managed to get it to boot now (still no drives) and get the logs (attached).

odin-diagnostics-20160915-1418.zip

Enabled rsync, got it to lock up again but nothing in the log, last entry is:

2016-09-17
07:29:34
Notice
ODIN
daemon
rpc.mountd
authenticated mount request from 10.0.0.3:842 for /mnt/user/ISOs (/mnt/user/ISOs)

 

Anyone know of any other logs I can check/way I can diagnose? Clutching at straws here.

Can you try upgrading your system to version 6.2?