Potential disk failing?

[jlw_4049]

Figured it would be best to start a new thread here since I was discussing it in this thread here and said I would post back. 

My server did not start to freeze this time but when I was rebuilding my parity it made it to about 88% and then then stopped due to the parity drive failing and saying Disabled.

 

When someone has the time would you please look at my logs and let me know what you think? It passes a short SMART test but skimming over the logs I see this quite frequently. 

Dec 20 00:23:08 jlw-unRaid kernel: md: disk0 write error, sector=24319852752

I checked the physical connections, although nothing has changed. I did put the server under a heavy load during the end of the parity check though, while I doubt it's power I guess that could be an option. It's a 10 year warranty PSU of pretty high quality, so I wouldn't think so. 

Thank you!

 

jlw-unraid-diagnostics-20231220-0036.zip

[itimpi]

It looks like it could be a controller issue as just before parity 0 started showing write errors you got.

Dec 20 00:23:03 jlw-unRaid kernel: mpt2sas_cm0: SAS host is non-operational !!!!
### [PREVIOUS LINE REPEATED 5 TIMES] ###
Dec 20 00:23:08 jlw-unRaid kernel: mpt2sas_cm0: _base_fault_reset_work: Running mpt3sas_dead_ioc thread success !!!!

and there is no SMART information for the parity drive in the diagnostics.

[JorgeB]

Make sure the HBA is well seated and sufficiently cooled, you can also try a new PCIe slot if available.

[jlw_4049]

This is the HBA card in my server. It's been in use since 2020 and I've had no issues. I pulled the side panels off, I pushed on the HBA card, but it was in there quite snugly didn't move or anything. I reseated the power/sata cables going to the parity drive, however I am still using the same cables. After this post last night I started an extended SMART test. Currently it's at 60%. 

When I did the diagnostics the drive was "disabled" or not reading I guess, so maybe that's why it didn't include the short SMART results. I attached the short SMART results, as I had tested them right before that. 

If the HBA was dropping off wouldn't it drop other drives off on it as well? Thanks for the replies so far. I'll update when the extended SMART run is completed. 

Additionally during this extended SMART run the CPU is at ~100% load so I could ensure no power issues (I'm encoding a huge queue of movies)

WDC_WD140EDGZ-11B2DA2_2CHEZJ8N-20231218-1804.txt

Edited December 20, 2023 by jlw_4049

[jlw_4049]

So it passed the extended test with 0 errors that I could see. 

I just took it off the array as parity, started, stopped, added it back to the array to start the parity sync. So I guess in about a day I'll see if things are sorted. Here is the extended smart test results. 

WDC_WD140EDGZ-11B2DA2_2CHEZJ8N-20231220-2236.txt

Edited December 21, 2023 by jlw_4049

[JorgeB]

Disk is OK for now, keep monitoring.

[jlw_4049]

Thanks for the response. Parity is complete with 0 errors. 

I'm using these cables on my LSI card 
https://www.amazon.com/gp/product/B012BPLYJC/ref=ppx_yo_dt_b_search_asin_title?ie=UTF8&th=1

They do connect to the HDD a little loose. I am wondering if I should just replace the cables since these are relatively cheap and loose to prevent these errors from happening again if it could be the issue. 

I'm wondering if over time during use the sata just got to a point where it wasn't connected good and caused all these issues. Is there any recommended brands for these type of cables? 

Edited December 22, 2023 by jlw_4049

[JorgeB]

The previous error was not cable related, but if they are not making a good contact they can also cause issues.

[jlw_4049]

The cables do work just don't fit snugly like a normal sata cable, perhaps it's a clipping issue on my set since they are pretty well reviewed. Do you have any further recommendations for me to do? 

[JorgeB]

Not really, keep monitoring.