One day after updating to unRaid 7, I began having issues. At first I noticed shares were no longer accessible through SMB. When checking the webUI I saw these shares were now not even listed. After checking the physical disks I could see that the folders and files were still there. After a reboot things would work again for a few hours and then it would happen again. At this point I noticed that my dockers would crash. Apparently something may be happening with my cache drive. Any share that had data on the cache drive would disappear from my shares list. Possibly a failing cache drive that by coincidence started failing after updating to 7?

tower-diagnostics-20250123_1517.zip

Syslog rotated so we cannot see the beginning of the issue, reboot and post new diags after array start.

Will do. I am testing cables, controller card, etc first to rule those out first.

It just happened again and I lost all but two shares in the webGUI. I grabbed the diagnostics immediately.

tower-diagnostics-20250124-1149.zip

Edited January 24, 20251 yr by kricker

Jan 24 11:46:44 TOWER kernel: ata10: hard resetting link
Jan 24 11:46:44 TOWER kernel: ata10: SATA link up 3.0 Gbps (SStatus 123 SControl 300)
Jan 24 11:46:44 TOWER kernel: ata10.00: model number mismatch 'PNY CS900 500GB SSD' != 'PS3111'
Jan 24 11:46:44 TOWER kernel: ata10.00: revalidation failed (errno=-19)
Jan 24 11:46:44 TOWER kernel: ata10: limiting SATA link speed to 1.5 Gbps
Jan 24 11:46:49 TOWER kernel: ata10: hard resetting link
Jan 24 11:46:50 TOWER kernel: ata10: SATA link up 1.5 Gbps (SStatus 113 SControl 310)
Jan 24 11:46:50 TOWER kernel: ata10.00: model number mismatch 'PNY CS900 500GB SSD' != 'PS3111'
Jan 24 11:46:50 TOWER kernel: ata10.00: revalidation failed (errno=-19)
Jan 24 11:46:50 TOWER kernel: ata10.00: disable device

 

Cache device dropped offline, replace cables and swap SATA ports and try again

That was after a cable swap and moving it from the onboard controller back to the promise SATA controller. I now cloned the Cache to a new SSD and swapped it out. I did see that limited speed warning on boot even with new SSD on the Promise SATA controller and it would not even show up in the webGUI when I booted. I shut down again, moved the Cache to the onboard controller and had no warning about the link speed this time. I was able to assign the new cache drive. So I will see how it goes now. If it is stable. I will move it to the Promise controller again with yet another new cable and see what happens.

 

Interesting when the Cache drops offline it takes the user shares with it. I guess that makes sense as those shares are all set to use the cache drive. Oddly I do not get any warnings/alerts in the webGUI and when it drops offline I still see it in the webGUI assigned in the cache pool. I have to to a lot of clicking back and forth through the tabs refreshing to see if it disappears. Several times it never did and always showed as assigned.

Edited January 24, 20251 yr by kricker

21 minutes ago, kricker said:

Interesting when the Cache drops offline it takes the user shares with it

That's normal, because the filesystem crashes:

 

Jan 24 11:46:55 TOWER kernel: XFS (sde1): Filesystem has been shut down due to log error (0x2).
Jan 24 11:46:55 TOWER kernel: XFS (sde1): Please unmount the filesystem and rectify the problem(s).

 

See if it's resolved now, if it still happens, recommend replacing the SSD.

As mentioned above. I have replaced the SSD with a new one. I cloned my old one to a new one and swapped it out. I am testing things now. I guess this was all just coincidence after upgrading to vs7. I have dropped back to 6.12.14 for now.

This should not be a software issue, at most it could be a compatibility issue with the kernel, maybe with the Promise controller, but would be extremely unlikely with an Intel Controller, or there would be a lot of users with issues.

I have two other drives on that promise controller. I don't recall them disappearing. So my hunch is it's just that SSD decided to finally take a dump.

Oddly enough it is that Promise controller now having issues with unRaid in V7. Everything was working fine with the cache drive connected to the onboard controller in 6.12.14. i upgraded back to 7 and again everything was fine. I ran it since my last post with no issues. Yesterday I shutdown and swapped it back to the Promise controller. Upon boot I saw the slow link warnings. I shut it back down. Swapped to yet another new cable but still using the Promise controller. Once it booted, the cache SSD was missing. I had to move it back to one of the onboard controller ports. 

 

Strangely, I still have drives 1 and 2 connected to that controller, but they are not having any issues that I am aware of.