Version 6.12.0

 

When I went into my office on Thursday (6/29) morning to start work I saw that my HP Microserver G8 had a red flashing status light.  After connecting to iLO to figure out what was going on, I found that it was a "power fault".  I did search for quite some time on the code thrown by this power fault and found nothing that specifically matched.  In fact most folks that had this same symptom were then unable to power on their server.  Mine powered right up.  Unraid stated there was an unclean shutdown and needed to do a parity check.  Approximately 15 hrs later, it completed with no errors.  I then started the array and it seemed fine.  Apps I have installed opened, I was able to interact like normal with the webUI, etc.  One weird thing was that removing an app (trying to free up some system resources by deleting unused or otherwise redundant apps.  Like a backup instance of Adguard Home, for example).  I received an error that stated it couldn't remove it because it didn't exist.

 

I thought that was a little strange, so I looked and saw there were some orphaned images (I forget the exact verbiage).  None were AGH though.  I removed all of those. 

 

This is the part that's fuzzy to me.   I should've written it down as it was happening but I panicked a bit.  Something happened with dockers, either the daemon stopped running/crashed or I did something.  I think part of this may have been I likely had way too many apps running with auto start up for the hardware I'm using.  Either way when I tried to start them again, they wouldn't budge.  I'd get the unraid logo squiggle but it would go back to 'array stopped'.  So I started in maintenance mode and that worked, but I wasn't exactly sure what to do with anything so I left it for a bit while I tried to research.  It was at this point I noticed that Disk 1 and 2 of my data array and both SSDs in my cache pool are showing the subject error message.  File system on HDD 1 & 2 show xfs.  The pool of two SSDs only shows on File system and that's btrfs.

 

lsblk results here:

 

Quote

NAME   MAJ:MIN RM   SIZE RO TYPE MOUNTPOINTS
loop0    7:0    0  58.8M  1 loop /lib
loop1    7:1    0 334.1M  1 loop /usr
sda      8:0    1  14.6G  0 disk
└─sda1   8:1    1  14.5G  0 part /boot
sdb      8:16   0  12.7T  0 disk
└─sdb1   8:17   0  12.7T  0 part
sdc      8:32   0  12.7T  0 disk
└─sdc1   8:33   0  12.7T  0 part
sdd      8:48   0  12.7T  0 disk
└─sdd1   8:49   0  12.7T  0 part
sde      8:64   0 931.5G  0 disk
└─sde1   8:65   0 931.5G  0 part
sdf      8:80   0 931.5G  0 disk
└─sdf1   8:81   0 931.5G  0 part /mnt/cache
md1p1    9:1    0  12.7T  0 md   /mnt/disk1
md2p1    9:2    0  12.7T  0 md   /mnt/disk2

 

Diagnostics attached.  I should mention that it wanted to do another parity check after I did a "clean shutdown".  But I didn't realize at the time the timing is so short by default so it did yet another unclean shutdown.

 

I did short SMART on disk 1 and it showed no errors.  Disk 2 is still running because I mis-clicked and selected extended by accident.  The webUI apparently didn't load in Firefox correctly so it looked lined up with the start for short test.  My mistake. 

 

I don't know what else to do.  I did see a few other posts with the unmountable symptoms and saw some moderators and other folks link stuff from the wiki, but the couple I tried were literally dead ends.  One landed at 404 page and the other went to the main wiki page.  I did see the banner that it's under construction, so that is likely why. 

 

If there is any suggestions I'm open to it.  If it's a done deal and I need to start over, also fine.  Annoying, but fine.  I had a decent chunk of stuffed backed up but it was still trying to do the initial backup when it had the power fault.  There is some critical data but thankfully I had it backed up in my iDrive and Backblaze B2 bucket (trying to get rid of iDrive and move to B2 entirely, but hadn't cut that cord yet, thank goodness.).

 

If anyone has any leads on the power fault that started this mess, please let me know.  I have the server connected to a APC UPS 1500VA on the surge/battery backup side before this happened.  And I even tested it by unceremoniously yoinking the power cord to the UPS from the wall and it did exactly what it's supposed to do (this was weeks ago, btw).  Probably a bit cavalier, but hey.  Still learning the ropes on this product.

borg-diagnostics-20230701-1946.zip

A reboot should fix it, also make sure you update to v6.12.2 and/or don't use Firefox, both should help prevent this issue.

I did reboot once after receiving the original unmountable messages.  No change there.  I've downloaded the OS update and I'm just waiting on the SMART test to finish and then I'll give it a reboot.  I'm assuming Chromium browsers fare better based on the advice to not use FF?

@JorgeB All good now!  Updated and rebooted after that extended SMART test and everything mounted and apps are loaded up.  

 

Given this scenario, is there anything else I should look into?  I've already scanned with the "fix common problems" plugin and it's good there too.   I also did a quick flash drive backup and I'll continue ensuring my data backups are good to go as well.

 

Thanks for the help!

9 hours ago, Father_Redbeard said:

is there anything else I should look into?

Everything should be fine now.