USERs List empty after crash

[dadarara]

was playing with UPS setup and had system crash while the Array was up and some VMs and DOCKER apps as well.

after reboot I wasnt able to login to SMB /AFP shares.

looked at the USERS tab and there was only the root user defined. 

with SSH access I see that indeed no users are defined in the linux system.

 

I have defined some users AGAIN  via WEB UI.  But after reboot the list got empty again.  

It doesnt keep any users I define permanent. 

 

the shares are intact as far as I can see. I can start VMS and Docker apps.

 

Diagnostics attached. (not that I know what it can help...)

 

tower-diagnostics-20170911-0437.zip

[dadarara]

anyone?

[dadarara]

I get the users deleted every time I make a reboot.  also the SSH access fails and I need to delete the known_hosts file in the remote computer.

 

anyone can direct me to how to investigate it ?

[dadarara]

?

[gubbgnutten]

On mobile so I can't check the diagnostics, but at least my first, second and third suspect would be the USB stick: Either file system corruption or something wonky preventing it from being mounted properly possibly causing a default config to be used, only stored in volatile memory.

 

Have you tried the Fix Common Problems plugin?

[dadarara]

common problems didnt show anything until yesterday.  found call traces.

if anyone can check the diagnostics please?

whats the best way to check and fix the USB stick ?

tower-diagnostics-20170918-1009.zip

[dadarara]

one more symptom

 

after reboot, I have my system TIME wrong.

I need to play with the NTP time definitions to make it right.

I add one more server to the list and apply. after that the time is synced.

 

something is really not right with my system.  I dont know what should I do, without total reinstall or messing with the file systems on my disks.

please help

 

 

[gubbgnutten]

One common step is to connect the USB stick to a Windows computer and check it there.

[dadarara]

connected to MAC.  checked the stick with DISK utility with firstAid.  nothing.

dont have windows..

Edited September 18, 2017 by dadarara

[gubbgnutten]

Could you post diagnostics from after the first aid attempt? Let's see if that got rid of the "Volume was not properly unmounted. Some data may be corrupt. Please run fsck." message.

 

The files FSCK0000.REC, FSCK0001.REC and FSCK0002.REC suggests previous corruption.

[dadarara]

attached the latest one

 

tower-diagnostics-20170919-1017.zip

I think there is still the error In the syslog

 

also, what is the suggested actions running the fsck on the USB stick ?

 

I run the bellow:

root@Tower:~# parted /dev/sdg 'print'

Model: Storage Xtreamer (scsi)

Disk /dev/sdg: 8099MB

Sector size (logical/physical): 512B/512B

Partition Table: msdos

Disk Flags: 

 

Number  Start   End     Size    Type     File system  Flags

 1      4194kB  8099MB  8095MB  primary  fat32        boot

 

root@Tower:~# fsck /dev/sdg1

fsck from util-linux 2.28.2

fsck.fat 3.0.28 (2015-05-16)

0x41: Dirty bit is set. Fs was not properly unmounted and some data may be corrupt.

1) Remove dirty bit

2) No action

? 2

There are differences between boot sector and its backup.

This is mostly harmless. Differences: (offset:original/backup)

  65:01/00

1) Copy original to backup

2) Copy backup to original

3) No action

? 3

/dev/sdg1: 3359 files, 135803/246912 clusters

root@Tower:~# 

Edited September 19, 2017 by dadarara

[gubbgnutten]

I would probably just recreate the USB stick, but that's me

 

Otherwise I would probably in addition to fsck also check the contents of the FSCK*.REC files and compare the rest of the files to the corresponding ones from a recent backup.

[dadarara]

but should I :

remove dirty bit ?

Copy original to backup or Copy backup to original

[dadarara]

?

[dadarara]

reviving this topic.

I am yet to decide to reinstall ALL the server.

 

can someone look at the log when the server starts?

1 - the NTP time service is not updating the time 

2 - the users definitions are still missing 

3 - SSH access always after reboot I need to delete the SSH.known user list from my remote pc. the identity of the unRAID is changing 

 

any ideas ?  hopping for some "easier" fix rather than reinstall.

 

 

tower-diagnostics-20171016-2232.zip

[dadarara]

anyone has an idea where to look?

[trurl]

Does sound like a problem with the flash device. Possibly it is dropping out after booting. Try a different USB port, preferably USB2.

[dadarara]

thanks

will probably reinstall everything. 

need to find the right instructions so not to delete anything 

 

[dadarara]

short update for posterity ..

reinstalled the unRAID on the same USB. 

done also MEMTEST 1 pass, just to make sure.  took long time for the 64GB I have

it kept all the data. 

after the process , the TIME settings and the user setting were sorted out OK.  no issues there.

after the process, 1st problem was that the  parity was with error. so I run the parity check. (10 long hours)

after that the party disk got disabled.  

had to remove the disk from configuration , stop start array and assign the disk again, then to run 10hours parity check again.

brrrrrrrrrrrrr

anyways, now it looks ok.

Also the responsiveness of the WEBGUI is much better. I think it became faster.