Parity errors again

[statikcat]

I posted a month or two ago about how I get disk and parity errors often with Unraid and after looking at the logs was told likely it is sata cable was not seating properly. Well I just logged in and saw parity had read errors on May 5th and it was trying to re-check parity since then and got stuck at 1.5%. So I stopped the array, unplugged sata and power to my drive and now parity check is chugging at 30-80MB/s (only 1% so far though). Could some of the kind experts here take a look at my log I attached? Also, my real time log is showing :

 

May 9 13:50:10 statiknas kernel: ata11: hard resetting link
May 9 13:50:20 statiknas kernel: ata11: softreset failed (1st FIS failed)
May 9 13:50:20 statiknas kernel: ata11: hard resetting link
May 9 13:50:21 statiknas kernel: ata11: SATA link up 3.0 Gbps (SStatus 123 SControl 320)
May 9 13:50:21 statiknas kernel: ata11.00: configured for UDMA/133
May 9 13:50:21 statiknas kernel: ata11: EH complete
May 9 13:50:33 statiknas kernel: ata11.00: exception Emask 0x10 SAct 0x7f0 SErr 0x9b0000 action 0xe frozen

 

These errors are common throughout my experience with Unraid regardless of drives or cables. It is always something going on like this. Is this.. another sata cable issue? My PC case has not even been bumped since my last post so I don't know how any cables are coming loose and they appear properly connected. Any ideas? Thanks!

statiknas-diagnostics-20220509-1647.zip

[statikcat]

Now my parity check is down to 3MB/s 33 days long..

I have 2 spare 8tb drives I can swap with it for testing if needed.

 

EDIT: Just used a different cable to a different port and restarted parity.. even worse at 800KB/s 80 days and counting..

Edited May 9, 2022 by statikcat

[Zonediver]

33 minutes ago, statikcat said:

Now my parity check is down to 3MB/s 33 days long..

I have 2 spare 8tb drives I can swap with it for testing if needed.

 

EDIT: Just used a different cable to a different port and restarted parity.. even worse at 800KB/s 80 days and counting..

 

What kind of hard drives are you using?

If you use Seagates, then... they are not really reliable...

[statikcat]

I am running all Seagate BarraCuda 8TB.  Similar drives in my Windows backup system never have any issues. Is it a general issue or dependent on file system?

Edited May 9, 2022 by statikcat

[Zonediver]

6 minutes ago, statikcat said:

I am running all Seagate BarraCuda 8TB....

 

Well... that's (maybe) the answer - sorry for that, but this drives are crap - just my opinion...

I used Seagats from 2000-2010 and almost all of them died... I avoid this manufacturer.

Edited May 9, 2022 by Zonediver

[statikcat]

What is recommended? WD?

[Zonediver]

10 minutes ago, statikcat said:

What is recommended? WD?

 

I am using WD-RED's since 2013 and lost only two drives over the years.

But generally the WDs are more reliable. I usually leave them in the system for 5 years and then replace them.

The new drives are Purple PROs because they have 5 years warranty and my server is only for Plex - so a native streaming-server.

The Purple PRO is ultrasilent, no vibrations and get only 32°C after 24h of constant writing (parity reconstruction).

These are impressive values...

Edited May 9, 2022 by Zonediver

[BRiT]

Seagate 8TB drives are perfectly fine. Many users run those drives without any issues.

[statikcat]

It looks like the ATA11 errors I posted above are the parity drive. If I click on the drive icon on the Main page under Identification I can see the errors from the log button in the upper right. The other 5 drives are clean. What do these errors indicate? I swapped cable and port and still get them (now on ata5). Is the drive bad? I guess the next step might be run an extended Smart test?

 

Interestingly, before discovering which drive the errors were for, I reseating all cables and rebooted unraid. Now the pairty drive is running at 100MB/s+ for over 10 minutes now. Strange! (the reset link and frozen warnings/errors still continue though..)

Edited May 9, 2022 by statikcat

[Zonediver]

5 minutes ago, statikcat said:

Interestingly, before discovering which drive the errors were for, I reseating all cables and rebooted unraid. Now the pairty drive is running at 100MB/s+ for over 10 minutes now. Strange!

 

Where is this drive connected? Mainboard? PCIe-Controller? Backplane?

Edited May 9, 2022 by Zonediver

[statikcat]

30 minutes ago, Zonediver said:

 

Where is this drive connected? Mainboard? PCIe-Controller? Backplane?

 

It was connected to a PCI controller (https://www.amazon.com/gp/product/B07ST9CPND/ref=ppx_yo_dt_b_search_asin_title?ie=UTF8&psc=1) along with many other drives. I moved it to the motherboard with a different cable (old first generation I7 mobo). However, it went really slow after using a different cable to the mobo as well. It was reseating all other cables and rebooting and now it is running fast again (although it has now slowed to 55MBs..) Hard to tell what is related.

 

I guess my plan for now, unless I get other advice, is to see if the parity can complete and then run some smart tests.

Edited May 9, 2022 by statikcat

[Zonediver]

6 minutes ago, statikcat said:

 

It was connected to a PCI controller (https://www.amazon.com/gp/product/B07ST9CPND/ref=ppx_yo_dt_b_search_asin_title?ie=UTF8&psc=1) along with many other drives. I moved it to the motherboard with a different cable (old first generation I7 mobo). However, it went really slow after using a different cable to the mobo as well. It was reseating all other cables and rebooting and now it is running fast again (although it has now slowed to 55MBs..) Hard to tell what is related.

 

I had a similar problem with one of my 3TB WD-Red's - very slow on some areas.

After some testing, it was clear that the platter coating was defective (weak sectors).

The disc had to be replaced. It had write-read rates of 3-5MB/s in some areas.

Edited May 9, 2022 by Zonediver

[JorgeB]

Looks like a power/connection problem, check/replace cables, both power and SATA.