Parity check slow - interface fatal error

[mattiapsu]

The parity check started last evening, and was only running at 1MB/s. I checked to see if anything was going on with the arrays, docker, vms, but nothing I could see there. I rebooted and attached two log files. One just after startup, one catching errors after parity started. Reading around I see solutions for these errors as simple as a bad SATA cable, or power issue, or worse. Looking at the disc logs, I can see that the error occurs on my parity disk. 

 

One other thing to add, I recently added a SATA pci card, and a second cache drive (running 1 cache drive for appdata, and a cache pool - 2drives for downloads). That was working fine for about 2 weeks from what I could tell, and pretty sure ran a successful parity check since then.

 

Before I start opening the box and switching cables, I wanted to get some expert advice. Thanks in advance.

oldmain-syslog-20220507-1416.zip oldmain-syslog-20220507-1148.zip

Edited May 7, 2022 by mattiapsu

[JorgeB]

Please post the diagnostics.

[mattiapsu]

I did confirm that I installed the new hardware and then ran a parity check that ran normally with no errors found. The only thing I did after the hardware/parity check and before this issue was remove some docker containers (photoprism, mariadb), delete their appdata folder, and then reinstall them. I wanted fresh containers to start over.

 

Thanks.

oldmain-diagnostics-20220508-1512.zip

[JorgeB]

Check/replace cables (both SATA and power) on parity and disk1.

[mattiapsu]

Thanks, will take me a couple of days to get to it. But I'll report back. 

[mattiapsu]

A couple of days ago I checked my cables, made sure everything was well connected. Ran a parity check after I closed back up, and it started slow, but returned to normal parity check speed. However, it noted over 50,000 errors once finished and said parity was valid. Errors did show up and are attached in syslog attachment. Started a check today to see how it would react, and started slow and picked up parity errors immediately, I canceled it without seeing if it sped up. Similar errors being thrown today as a couple of days ago.

 

Could it be a power supply issue? I'm using a 400W supply (started with modest ambitions, but I've added more hardware). I can change out data cables as well to check those, but it may be very apparent to you that the PSU is undersized.

 

Quick snapshot of hardware

- 2 sticks RAM

- 1080 Ti GPU used in VM CCTV only

- USB 3.0 PCIe card

- Sata PCIe expansion card

- 3 HDD (2 data, 1 parity)

- 2 SSD (1 appdata cache, 1 cache pool)

- 2 USB HDD (old laptop drives) - 1 as unassigned drive, 1 cache pool

 

 

oldmain-syslog-20220515-1852.zip

[JorgeB]

There are still ATA errors, if you didn't replace cables at least swap them around to see if the problem follows the cables.

[mattiapsu]

I swapped sata cables around (3 way swap between 3 HDD). The ATA error stuck with the parity drive but moved from ata1 to ata2. I didn't mess with any power cables at this point.

 

oldmain-syslog-20220524-0241.zip

[JorgeB]

15 minutes ago, mattiapsu said:

The ATA error stuck with the parity drive

That rules out the SATA port and SATA cable, leaves the power cable or the drive as the possible issues.

[mattiapsu]

I finally took the time to pull the box down and mess with power cables.

1) plugged new power cable into drive, started up.... showed as drive missing; opened back up

2) plugged old power cable back into drive, started up... showed up with zero errors and running parity check happily

 

I'll monitor, but does this sound like,

1) a random occurrence, slightly bad cable connection

2) a power cable or PSU issue

3) a drive issue?

 

Appreciate your experience here. I don't have spare PSU or drives around, so depending on your thoughts may get some backups. 

 

Thanks

 

[JorgeB]

Difficult to say for now, see how it goes in the future.