Random freezing when making writes

[dbuhr]

This has been going on for a while, when downloading/extracting nzbs, the entire server seems to "lock up" (nothing in syslog). It is not consistent and usually recovers in a minute or so.  However, I recently added Roon which will cause the entire server lock up every few minutes when it analyzes my library to the point it itself crashes.

 

Any ideas?

 

 

diagnostics.zip

[Squid]

 Is the share the downloads are happening on name "Downloads"?  That share exists on every disk.  Do your drives spin down?  Most controllers if they have to spin up another drive (if the OS is looking for a file or something) wind up pausing all I/O traffic until that drive is spun up.

[dbuhr]

The Downloads share is where the action occurs for nzbs, but Roon was writing to appdata, so this could be 2 separate issues i'm conflating. 

 

I did a bit more digging, and I'm starting to think something is wrong with my cache array.  I fired up a VM that was running Roon previously and realized it is showing out-of-space (domains share I believe) 99.8% used of 8GB.  I also got Roon running on Docker, pointing to a share that only runs on the array, no cache, and it seems to be working fine now.

 

I did follow this article, however, neither drive shows full and the rebalance didn't help.  I did plan on adding more space, maybe my old 250GB SSDs aren't sufficient, even though it seems like they should be?

Label: none  uuid: 169dda69-6f92-4350-b51a-d9c7d5b9fbdf
        Total devices 2 FS bytes used 45.48GiB
        devid    1 size 232.89GiB used 73.03GiB path /dev/sdf1
        devid    2 size 232.89GiB used 73.03GiB path /dev/sde1

Data, RAID1: total=72.00GiB, used=45.36GiB
System, RAID1: total=32.00MiB, used=16.00KiB
Metadata, RAID1: total=1.00GiB, used=119.44MiB
GlobalReserve, single: total=39.62MiB, used=0.00B