Lots of Parity Errors

[Joe]

After running unRaid since v4 with no parity issues or other problems I decided it was time for some maintenance and upgrades.  All my drives were still reiserfs and I was getting tight on room.  I bought 2 14 TB Toshiba drives off eBay and ran them through the pre-clear to test them.  One failed and sent to Toshiba (where it is stuck because they  have no replacements).  The "good" one was to be my new parity drive.  I did a parity check with 0 errors.  I powered down the server, replaced my Hitachi 4TB with my 14TB Toshiba and let it rebuild.  All good.  Took old parity and pre-cleared and formated it to XFS and added it to the array.  I started moving data to empty drives using MC and Krusader, then reformatting from reiserfs to XFS.  I am getting parity errors in the high hundreds while moving files.  I am wondering if anyone has any ideas and what I should do.  I am attaching system log and screen shot of setup.  

Thanks

Joe 

tower 2 system log.txt

[JorgeB]

Please post the complete diagnostics but looking at the syslog it appears parity is failing.

[Joe]

Attached

tower2-diagnostics-20200509-1439.zip

[JorgeB]

Some weird SMART errors on the parity disk, you should run an extended SMART test.

[Joe]

Any idea how long an extended SMART takes on a 14TB drive?  I think it is bad as I am getting 

Unraid Parity disk SMART health [197]: 13-05-2020 22:07

Warning [TOWER2] - current pending sector is 160

[JorgeB]

3 hours ago, Joe said:

Any idea how long an extended SMART takes on a 14TB drive? 

SMART report will show aprox time, e.g.:

 

Extended self-test routine recommended polling time: 1128minutes.

 

 

[Joe]

Mine is going 9 hours and only 40%.  Probably be a total of 24 hours to finish

[JorgeB]

17 minutes ago, Joe said:

Probably be a total of 24 hours to finish

What's the polling time for that drive?

[Joe]

Where do I check polling time?

Thanks

[JorgeB]

7 hours ago, johnnie.black said:

SMART report will show aprox time

 

[Joe]

I have attached the smart report.  I do not see polling time.  The drive is going bad as I have 960 pending sectors.  I bought 2 of these Toshiba 14TB drives (eBay great price, bad drives but under warranty) and have already RMA'd one drive, should have it back Monday.  They don't have replacements in stock so they are sending me a 4kn drive back and I hope it will run on my LSI 9211.  I just upgraded the firmware to P20 and saw a uTube vid that says it will but read other places it will not.  Nothing like doing it yourself to find out.   I thought if it pre-cleared without issue it would have been fine so I put it in as my parity drive and that is when the fun started.

Since I have run this machine for about 7 years without any issues, maintenance or upgrades, I figured I would buy some large drives and finally do some consolidation and convert my disks from RFS to XFS.  I was even thinking of moving from my FX-8320 to a Ryzen 7 to cut energy and add some threads to do more VMs (I'll wait until I sort out all my disk issues).   

Thanks

Joe

tower2-smart-20200514-1024.zip

[JorgeB]

Extended self-test routine
recommended polling time:      (1382) minutes.

 

[Joe]

Thanks - about 23 hours.  That's a long test.  I will do that on the replacement drive when it comes in, and if it passes, I will pre-clear, then move to parity, rebuild, then run another check.  That is about a day for each.  It is amazing how much time this takes.  Then when I get the other 14TB drive back, the same, except I will replace a drive, rebuild from parity and move data from 2 other drive to it also.  

[JorgeB]

3 minutes ago, Joe said:

about 23 hours.  That's a long test. 

Yep, it's a complete surface scan, the larger the disk, the longer it will take, that disk isn't looking very good, but wait for the test to finish.