Hi, 

 

I have 4 8TB hdd, one of them is the parity drive. 

The last parity check is a week ago. There's been lots of writing recently. And I upgraded the system from 6.8.3 to 6.9.0-rc1.

I've set the parity check to run once a week. 

Yesterday I upgraded the system to 6.9.0-rc1

Last night's parity check reports 332 errors, it seems to be too many, is it normal?

Drives report no error, mostly less than one month old. 

RAM has been tested via memtest for 18 hours without error before the setup.

 

What's the recommended frequency to run parity check?

 

Dec 13 07:47:23 Tower kernel: md: recovery thread: P corrected, sector=8059212888
Dec 13 07:47:23 Tower kernel: md: recovery thread: P corrected, sector=8059212896
Dec 13 07:47:23 Tower kernel: md: recovery thread: P corrected, sector=8059212904
Dec 13 07:47:23 Tower kernel: md: recovery thread: P corrected, sector=8059212912
Dec 13 07:47:23 Tower kernel: md: recovery thread: P corrected, sector=8059212920
Dec 13 07:47:23 Tower kernel: md: recovery thread: P corrected, sector=8059212928
Dec 13 07:47:23 Tower kernel: md: recovery thread: P corrected, sector=8059212936
Dec 13 07:47:23 Tower kernel: md: recovery thread: P corrected, sector=8059212944
Dec 13 07:47:23 Tower kernel: md: recovery thread: P corrected, sector=8059212952
Dec 13 07:47:23 Tower kernel: md: recovery thread: P corrected, sector=8059215776
Dec 13 07:47:23 Tower kernel: md: recovery thread: P corrected, sector=8059215784
Dec 13 07:47:23 Tower kernel: md: recovery thread: P corrected, sector=8059215792
Dec 13 07:47:23 Tower kernel: md: recovery thread: P corrected, sector=8059215800
Dec 13 07:47:23 Tower kernel: md: recovery thread: P corrected, sector=8059215808
Dec 13 07:47:23 Tower kernel: md: recovery thread: P corrected, sector=8059215816
Dec 13 07:47:23 Tower kernel: md: recovery thread: P corrected, sector=8059215824
Dec 13 07:47:23 Tower kernel: md: recovery thread: P corrected, sector=8059215832
Dec 13 07:47:23 Tower kernel: md: recovery thread: P corrected, sector=8059215840
Dec 13 07:47:23 Tower kernel: md: recovery thread: P corrected, sector=8059215848
Dec 13 07:47:23 Tower kernel: md: recovery thread: P corrected, sector=8059215856
Dec 13 07:47:23 Tower kernel: md: recovery thread: P corrected, sector=8059215864
Dec 13 07:47:23 Tower kernel: md: recovery thread: P corrected, sector=8059215872
Dec 13 07:47:23 Tower kernel: md: recovery thread: P corrected, sector=8059215880
Dec 13 07:47:23 Tower kernel: md: recovery thread: P corrected, sector=8059215888
Dec 13 07:47:23 Tower kernel: md: recovery thread: P corrected, sector=8059215896
Dec 13 07:47:23 Tower kernel: md: recovery thread: P corrected, sector=8059215904
Dec 13 07:47:23 Tower kernel: md: recovery thread: P corrected, sector=8059215912
Dec 13 07:47:23 Tower kernel: md: recovery thread: P corrected, sector=8059215920
Dec 13 07:47:23 Tower kernel: md: recovery thread: P corrected, sector=8059215928
Dec 13 07:47:23 Tower kernel: md: recovery thread: P corrected, sector=8059215936
Dec 13 07:47:23 Tower kernel: md: recovery thread: P corrected, sector=8059215944
Dec 13 07:47:23 Tower kernel: md: recovery thread: P corrected, sector=8059215952
Dec 13 07:47:23 Tower kernel: md: recovery thread: P corrected, sector=8059215960
Dec 13 07:47:23 Tower kernel: md: recovery thread: P corrected, sector=8059215968
Dec 13 07:47:23 Tower kernel: md: recovery thread: P corrected, sector=8059215976
Dec 13 07:47:23 Tower kernel: md: recovery thread: P corrected, sector=8059215984
Dec 13 07:47:23 Tower kernel: md: recovery thread: P corrected, sector=8059215992
Dec 13 07:47:23 Tower kernel: md: recovery thread: P corrected, sector=8059216000
Dec 13 07:47:23 Tower kernel: md: recovery thread: P corrected, sector=8059216008
Dec 13 07:47:23 Tower kernel: md: recovery thread: P corrected, sector=8059216016
Dec 13 07:47:23 Tower kernel: md: recovery thread: P corrected, sector=8059216024
Dec 13 07:47:23 Tower kernel: md: recovery thread: P corrected, sector=8059216032
Dec 13 07:47:23 Tower kernel: md: recovery thread: P corrected, sector=8059216040
Dec 13 07:47:23 Tower kernel: md: recovery thread: P corrected, sector=8059216048
Dec 13 07:47:23 Tower kernel: md: recovery thread: P corrected, sector=8059216056
Dec 13 07:47:23 Tower kernel: md: recovery thread: P corrected, sector=8059216064
Dec 13 07:47:23 Tower kernel: md: recovery thread: P corrected, sector=8059216072
Dec 13 07:47:23 Tower kernel: md: recovery thread: P corrected, sector=8059216080
Dec 13 07:47:23 Tower kernel: md: recovery thread: P corrected, sector=8059216088
Dec 13 07:47:23 Tower kernel: md: recovery thread: P corrected, sector=8059216096
Dec 13 07:47:23 Tower kernel: md: recovery thread: P corrected, sector=8059216104
Dec 13 07:47:23 Tower kernel: md: recovery thread: P corrected, sector=8059216112
Dec 13 07:47:23 Tower kernel: md: recovery thread: P corrected, sector=8059216120
Dec 13 07:47:23 Tower kernel: md: recovery thread: P corrected, sector=8059216128
Dec 13 07:47:23 Tower kernel: md: recovery thread: P corrected, sector=8059216136
Dec 13 07:47:23 Tower kernel: md: recovery thread: P corrected, sector=8059216144
Dec 13 07:47:23 Tower kernel: md: recovery thread: P corrected, sector=8059216152
Dec 13 07:47:23 Tower kernel: md: recovery thread: P corrected, sector=8059216160
Dec 13 07:47:23 Tower kernel: md: recovery thread: P corrected, sector=8059216168
Dec 13 07:47:23 Tower kernel: md: recovery thread: P corrected, sector=8059216176
Dec 13 07:47:23 Tower kernel: md: recovery thread: P corrected, sector=8059216184
Dec 13 07:47:23 Tower kernel: md: recovery thread: P corrected, sector=8059216192
Dec 13 07:47:23 Tower kernel: md: recovery thread: P corrected, sector=8059216200
Dec 13 07:47:23 Tower kernel: md: recovery thread: P corrected, sector=8059216208
Dec 13 07:47:23 Tower kernel: md: recovery thread: P corrected, sector=8059216216
Dec 13 07:47:23 Tower kernel: md: recovery thread: P corrected, sector=8059216224
Dec 13 07:47:23 Tower kernel: md: recovery thread: P corrected, sector=8059216232
Dec 13 07:47:23 Tower kernel: md: recovery thread: P corrected, sector=8059216240
Dec 13 07:47:23 Tower kernel: md: recovery thread: P corrected, sector=8059216248
Dec 13 07:47:23 Tower kernel: md: recovery thread: P corrected, sector=8059216256
Dec 13 07:47:23 Tower kernel: md: recovery thread: P corrected, sector=8059216264
Dec 13 07:47:23 Tower kernel: md: recovery thread: P corrected, sector=8059216272
Dec 13 07:47:23 Tower kernel: md: recovery thread: P corrected, sector=8059216280
Dec 13 07:47:23 Tower kernel: md: recovery thread: P corrected, sector=8059216288
Dec 13 07:47:23 Tower kernel: md: recovery thread: P corrected, sector=8059216296
Dec 13 07:47:23 Tower kernel: md: recovery thread: P corrected, sector=8059216304
Dec 13 07:47:23 Tower kernel: md: recovery thread: P corrected, sector=8059216312
Dec 13 07:47:23 Tower kernel: md: recovery thread: P corrected, sector=8059216320
Dec 13 07:47:23 Tower kernel: md: recovery thread: P corrected, sector=8059216328
Dec 13 07:47:23 Tower kernel: md: recovery thread: P corrected, sector=8059216336
Dec 13 07:47:23 Tower kernel: md: recovery thread: P corrected, sector=8059216344
Dec 13 07:47:23 Tower kernel: md: recovery thread: P corrected, sector=8059216352
Dec 13 07:47:23 Tower kernel: md: recovery thread: P corrected, sector=8059216360
Dec 13 07:47:23 Tower kernel: md: recovery thread: P corrected, sector=8059216368
Dec 13 07:47:23 Tower kernel: md: recovery thread: P corrected, sector=8059216376
Dec 13 07:47:23 Tower kernel: md: recovery thread: P corrected, sector=8059216384
Dec 13 07:47:23 Tower kernel: md: recovery thread: P corrected, sector=8059216392
Dec 13 07:47:23 Tower kernel: md: recovery thread: P corrected, sector=8059216400
Dec 13 07:47:23 Tower kernel: md: recovery thread: P corrected, sector=8059216408
Dec 13 07:47:23 Tower kernel: md: recovery thread: P corrected, sector=8059216416
Dec 13 07:47:23 Tower kernel: md: recovery thread: P corrected, sector=8059216424
Dec 13 07:47:23 Tower kernel: md: recovery thread: P corrected, sector=8059216432
Dec 13 07:47:23 Tower kernel: md: recovery thread: P corrected, sector=8059216440
Dec 13 07:47:23 Tower kernel: md: recovery thread: P corrected, sector=8059216448
Dec 13 07:47:23 Tower kernel: md: recovery thread: P corrected, sector=8059216456
Dec 13 07:47:23 Tower kernel: md: recovery thread: P corrected, sector=8059216464
Dec 13 07:47:23 Tower kernel: md: recovery thread: P corrected, sector=8059216472
Dec 13 07:47:23 Tower kernel: md: recovery thread: P corrected, sector=8059216480
Dec 13 07:47:23 Tower kernel: md: recovery thread: P corrected, sector=8059216488
Dec 13 07:47:23 Tower kernel: md: recovery thread: P corrected, sector=8059216496
Dec 13 07:47:23 Tower kernel: md: recovery thread: stopped logging

 

Edited December 13, 20205 yr by rampage

1 hour ago, rampage said:

Last night's parity check reports 332 errors, it seems to be too many, is it normal?

1 sync error is too many, any unclean shutdowns since last check?

When I tested this setup at the beginning, I only use two 8TB hdd as pool , and no parity drive, at that time there's unlcean shutdowns. 

 

But later when I was happy with everything, I added the parity drive and one more 8TB, since that time there's no unclean shutdowns.

I have written about 15TB into the 24TB pool, mostly through USB 3.0 unassigned drive , rsync or cp through the shell.

If there were no unclean shutdowns there's likely a hardware issue, you should post the diags so we can look at the hardware used.

Thanks for taking time to help me. 

 

Attached is the diagnostics. I also clicked upload hardware profile.

 

It is possible there's some hardware problem, it's an old computer. 

I can't tell what could be the issue.

tower-diagnostics-20201213-2110.zip

SMART data for your disks looks OK, doesn't completely rule them out but I would be look at simple fixes first try changing SATA cables and or SATA ports for your disk. I had a similar issue two weeks ago and for me it was my H310 controller card getting too hot but I suggest starting with the simple things different cables and SATA ports first and stay away from the Mavell SATA controller on that X58A-UD3R board too for UNRAID if you can.

 

Switch out your SATA cables, ports too if you can and run a parity check with corrections, once that's done run another (with or without corrections) if you get parity errors after the second run when the errors are supposed to be corrected then time to dig deeper.

Edited December 13, 20205 yr by Greg-Mega

You should also run memtest, bad RAM is the number 1 cause for sync errors.

Thank you, will try the suggestions