I've not had a successful parity check in 3 months, and it's been abysmally slow for 6 months..  The speed is 1 MB/sec or less at points, and its estimated time bounces between 60-110+ days, when it used to normally take about 17 hours, give or take.  The speed is equivalent to the write speed on the parity drives, as it is rebuilding.  I ran the disk speed docker and it was giving me errors on a 1TB seagate drive I had in there, but had nothing on it.  So I moved those folders to another drive, and removed it from the array thinking that might be the issue.  But parity is rebuilding at the same rate as it was checking before.

 

Strangely, the read speed of all my drives is way up there from 30+ MB/sec to hundreds of MB/sec.  The network dashboard is also showing tons of traffic exiting the network interface, yet my network controller says there is barely any traffic coming and going to it at all, measured in a few bps either up or down.  Docker and VM's are disabled as well, not that that matters, as both Docker and VM's are on the Nvme drives and not on the array at all.

 

Temps on all drives are between 79 - 91 degrees F depending on the drive.

 

Running Unraid 6.12.13

It has an i9-13900F processor, with 64GB RAM, Motherboard ASUSTek Prime H770-Plus D4

2 8TB parity drives

4 4TB data drives (was going to add in more until this parity problem started, so that's why there are no 8TB data drives at the moment)

 

 

Anyone have any ideas of why parity has slowed down so much.  The last time it ran with normal speed was back in March.

 

Thanks,

Scott

server-diagnostics-20241008-1818.zip

Edited October 8, 20241 yr by Aquenon

You are likely to get more informed feedback if you attach your system's diagnostics zip file to your next post in this thread.  
It is always a good idea to do this to allow us to see the current state of your system and so we can see logs.

Thanks, I knew I was forgetting something.

 

Those 8TB have been in there for a year or so and were running fine.  By the time I thought I needed to start thinking about adding more in, this parity issue was happening.

I attached the diagnostics to the original post.  I’m also getting these notices in mail as it rebuilds parity:

 

Event: Unraid Status
Subject: Notice [SERVER] - array health report [FAIL]
Description: Array has 10 disks (including parity & pools)
Importance: alert

Parity - WDC_WD80EFBX-68AZZN0_VGGSV5BG (sdf) - active 91 F [DISK INVALID]
Parity 2 - WDC_WD80EFBX-68AZZN0_VGGX0XXK (sdi) - active 90 F [DISK INVALID]
Disk 1 - WDC_WD40EFRX-68N32N0_WD-WCC7K3HRJUPK (sdc) - active 84 F [OK]
Disk 2 - WDC_WD40EFRX-68N32N0_WD-WCC7K1EJS37C (sdj) - active 86 F [OK]
Disk 3 - WDC_WD40EFZX-68AWUN0_WD-WXA2DA164VE6 (sde) - active 81 F [OK]
Disk 4 - WDC_WD40EZAZ-00SF3B0_WD-WX32DC08ETF5 (sdb) - active 82 F [OK]
Cache - WDC_WDS500G1R0A-68A4W0_21436W801671 (sdg) - active 88 F [OK]
Cache 2 - WDC_WDS500G1R0A-68A4W0_214513800434 (sdh) - active 91 F [OK]
Nvme - Samsung_SSD_970_EVO_250GB_S465NX0KA08969Z (nvme0n1) - active 88 F [OK]
Nvme 2 - Samsung_SSD_970_EVO_250GB_S465NB0K921298F (nvme1n1) - active 84 F [OK]

Parity-Sync in progress.
Total size: 8 TB
Elapsed time: 6 hours, 43 minutes
Current position: 415 GB (5.2 %)
Estimated speed: 1.3 MB/sec
Estimated finish: 65 days, 22 hours, 59 minutes
Sync errors corrected: 0

Edited October 8, 20241 yr by Aquenon

There's something reading from multiple disks, stop all other activity.

Thank you!  This is the second version of this reply, because I’d already shut down Docker and VM, which shouldn’t have made a difference anyway as they are entirely on my NVMe drive pool.  I had deleted half my plugins already.  I couldn’t figure out what else it could be.  But your reply reminded me that I hadn’t deleted the rest of my plugins.  At first thought, I wanted to use glances, but that would’ve meant re-enabling docker.  I knew I could get some info from the top plug-in, and when I ran it, I saw the issue.  I had multiple ssh processes running.  Turns out it was the Dynamix File Integrity plug in.  Deleted that plug-in, and suddenly I’m back to under a day to complete.  So I came back here, erased what I had written, and started over.

 

All that seems to be running now is unraidd0 mostly, but also a little bit from shfs, mdrecoveryd, and kworker/1:1H-kblockd.  There’s a few other processes that show up briefly, but those are the only ones consistently active at ~1 to ~2% except for unraidd0 which is running around 10%.

 

I like to think I would have caught that eventually, but it’s been months so maybe not as I had thought I had already turned everything off.  So thank you so much for the kick!