Consistent Writes to Cache Drive - 6.12.8

[strictlyforcatpics]

Hi everyone, fairly new to Unraid but just upgraded my hardware and have a question about some activity I noticed.

 

My system is exclusively used for Plex for the time being. Prior to yesterday my setup was an HP EliteDesk mini PC with an m.2 Cache drive and my Pool consisted of a 2.5" WD 500GB internal drive and then 2 Seagate 4Tb in an Orico USB Drive enclosure, Parity was a WD 4tb drive also in the USB drive enclosure. It seemingly worked well except after a shutdown I would need to rebuild parity as the disks would show as incorrect. I always knew that setup would be temporary so it wasn't a huge deal to me.

 

Over the past few days I built a new system that would allow me to connect all my drives directly through SATA. At first I decided to forgo the cache drive and the 500GB, as I never really used it in the other system, I had set my Plex share to write directly to the array for whatever reason. After getting everything back up and running, creating a new config with just the 2 Seagate pool drives and WD parity. All went well and Plex was back up and running normally. Though, I noticed that even when seemingly doing nothing I was getting random writes on drive 1 and parity and that if I spun the disks down Disk 1 would kick right back up. I stopped Docker, deleted Prometheus and Grafana App data folders (I decided to try it out, but ultimately didn't feel I would use it beyond its initial WOW factor), and was still getting those random writes. I did some research and found a lot of Unraid community/reddit user had their app data and system shares on a cache drive, so I figured maybe that was it, and that if I installed a cache and moved those shares to it, those writes would go away. Again my knowledge on this is very surface level and looking back I don't know why it would really make a difference. So I grabbed a new SSD (sata) and installed it (I'm going to use that HP for something else and wanted to keep that m.2 in there), set appdata and system primary storage to Cache, stopped docker, invoked Mover, waited until it seemingly finished and then started everything back up. Still getting those random writes. I know Plex can do a lot of background work, so I stopped that, no change.  I've run also run tips and tweaks/common fixes and have checked Active Streams and have still come up dry.

 

Is this normal for these shares? If not does anyone know what a fix could be?

 

tower-diagnostics-20240324-1832.zip

Edited March 24 by strictlyforcatpics

[itimpi]

It is perfectly normal for the drive(s) holding the 'system' share in particular to get regular activity as that is where the docker sub-system constantly has files open all the time it is running (which was what would have kept disk1+parity constantly spun up when that share was on disk1.

 

In addition the same consideration applies to the 'appdata' share any time you have any container running which is configured to use it.

 

In summary from your description you are just seeing the normal activity generated by running docker containers.   The whole idea is to keep such writes from going to the main array both for performance reasons and to avoid keeping array disks spun up.

[strictlyforcatpics]

Awesome, thanks for the help!