Slightly strange one; running a parity check since last night and noticed that the speed was much lower than usual - averaging around 70-75 mb/s insterad of 140-150 (about half). Stopped docker and that didn't make any difference (didn't think it would given what was running). Having looked at some posts online, I ran diagnostics - and immediately speed jumped back up to 150 Mb/s.

Does the diagnostics process restart a service, anything like that? Kicked something back into life? Not a particularly serious problem, just seemed weird. I've attached the diagnostics archive.

Thanks,

tower-diagnostics-20260303-1853.zip

First time I hear of that; maybe there was some other array activity going on that stopped at the same time?

Don't think so; i'd stopped the docker service a few minutes before but that didn't do anything. It had been running at the slow speed for about 20 hours, then in the space of 30 seconds between running the diagnostics and going back to 'Main', went straight to normal speed. Hard to imagine that was a complete coincidence although I guess not impossible. Server has been up for a few months so that was why I was wondering about some process restarting as part of the diagnostics. That might just be from being more used to Windows though......

Were the disks spun down when you started the check?

looking at the syslog, probably parity + 1 drive were spun up, rest down.

Edited March 4Mar 4 by Shamutanti
typo

Next time, try spinning up the drives before starting to see if it makes a difference. I've seen that before, where if you wake them up for a check, they may show degraded performance, and possibly getting the SMART reports for those made a difference.

Thanks Jorge, I'll give that a try.