UnRaid locking up

[Red_Stag]

Dear All,

 

Am running into issues with my unraid. Feel like I am chasing my tail each time I come across an error I address only to find another error and another. Have replaced parity with brand new drive in response to errors. Have replaced two other drives in response to errors but the errors just keep coming. Can't make out what is going on. Would very much appreciate a little guidance from someone more experienced than myself. I have about 2 or 3 months worth of Unraid experience.  Until I get more experience I am assuming that I have done something wrong.

 

Please advise if you are able to assist?

 

Red Stag

henderson-diagnostics-20231103-0703.zip

[kernelpaniced]

If you have a lot of plugins installed then try to boot into safe mode and see if your system is more stable. I had issues with a plugin causing my server to lock up and found it by doing this.

[Hammy Havoc]

11 hours ago, kernelpaniced said:

If you have a lot of plugins installed then try to boot into safe mode and see if your system is more stable. I had issues with a plugin causing my server to lock up and found it by doing this.

 

Any particular plugins that you felt were responsible?

[Red_Stag]

Thank you for the feedback. I never thought to consider plugins because in my humble opinion I thought that I had relatively few and only installed on an as needs basis.  The list of plugins installed is as follows:

 

1. community.applications.plg - 2023.09.30  (Up to date)
2. fix.common.problems.plg - 2023.10.08a  (Up to date)
3. unassigned.devices.plg - 2023.10.08  (Up to date)
4. unassigned.devices-plus.plg - 2023.08.14  (Up to date)
5. unassigned.devices.preclear.plg - 2023.08.14  (Up to date)
6. unbalance.plg - v2021.04.21  (Up to date)
7. unRAIDServer.plg - 6.12.4
8. user.scripts.plg - 2023.09.30  (Up to date)

I have considered a few different things to date:

 

Used a brand new PC so am inclined to assume that this is not the cause until proven otherwise (i7-12700, Gigabyte Z690, 32GB Ram)

Used brand new disks so am inclined to assume that this is not the cause until proven otherwise (Seagate 4TB CMR disks)

Used brand new cables so am inclined to assume that this is not the cause until proven otherwise

Used brand new PCIe To SFF 8087 MIniSAS SATA3.0 8 Ports Expansion Adapter ASM1166 Chip 20Gbps  (x2)

 

Feels like I am chasing my own tail, tried different CPU's, PCIe cards, PC's, hdd's. Errors manifest in different ways each time eg Unraid machine locks up and is not responsive or corrupted xfs filesystem or disk failure (even for a brand new disk), or parity failure. I have had all these errors since I started using Unraid on 24 August. I havent considered if it relates to the USB stick but the stick is relatively new (<2 years old with little use).  I have scoured the diagnostics files which are attached and I cant see anything which stands out as an obvious error.

 

Have put the above PCIe card into 4 PC's and each PC gives me different error messages at boot time. So I bought a second card from a different supplier and again, get different error messages on each PC and different error messages from the first PCIe card. This is my key reason for gravitating towards issues with the PCIe card.  I just dont feel like I have much else to point a finger at.

 

When I get a chance I will try the safe mode option and see what happens and reply.

 

The only other thing which occurs to me is the issue of hyperthreading, is it absolutely essential that the CPU has hyperthreading? Of the 4 PC's I have tried 2 have hyperthreading and 2 dont and the experience has been really challenging irrespective of the CPU.  Perhaps someone might be able to comment on this aspect please?  WOudl very much welcome any feedback and or comments please?

 

Many thanks

Red Stag

[JorgeB]

Enable the syslog server and post that after a crash.

 

[Red_Stag]

Thank you for the guidance JorgeB. I have attempted to do as asked. In the attached screenshot you will see that I have enabled the syslog.  Unfortunately since committing these settings I do not see any evidence of a log file created in the "appdata" folder. After this I enabled "mirror syslog to flash" in the pursuit of a log file. This worked and I attach the syslog file.  Perhaps I have a permissions issue which prevents me from writing files to the "appdata" folder?

 

I had to rebuild the UnRaid server again (took over 16 hours) so I created a new config file and it now only contains brand new 4TB CMR disks. I attach the updated diagnostics file as well.

 

I understand that the attached syslog file may not be of limited assistance because it is not a syslog file after a crash. I expect to have to wait a few days for a crash. But in the interim I provide the information that I do have and ask if there is some way that I can get the syslog to write to an array folder instead of to the flash drive please?

 

I have scanned the syslog file and the only thing that I observe which seems prospectively questionable is the repeated "Device or resource busy" notice.

 

Many thanks for any guidance offered

henderson-diagnostics-20231108-0729.zip syslog(1)

[itimpi]

You will be getting a file in the ‘log’ folder on the flash drive because you have the mirror to flash option set.   To get a file written to the ‘appdata’ share you need to put the IP address of your server in as the ‘remote server’ field as mentioned in the syslog server link.

[Red_Stag]

Dear itimpi, I clearly misinterpreted that, thank you for clarifying. I confirm that the log is created in the "appdata" folder now and I simply await a system crash now and will be back in touch with the syslog file.This may take a few days.  Thank you again

[Red_Stag]

Dear All,

 

As requested I have waited until unRaid got into an unresponsive state and I now attach the log and diagnostics files. I also attach photos showing the CPU activity. Once unRaid got into this state I decided to stop the array. Stopping the array took 11 minutes to complete. Five minutes prior to the unresponsive state I had completed a backup of unRaid data to unassigned devices using a luckybackup docker. Unclear to me if these are related as the time lapse seems long enough to seperate the two IMHO. 

 

Photo of Htop doesn't point the finger at anything that I can determine.

 

Would be very pleased if I could get some pointers please?

 

Many tahks

Red Stag

 

henderson-diagnostics-20231120-1945.zip henderson-syslog-20231120-0945.zip