Hello! I recently did a number of upgrades to my server and am out of other options. I have spent the better part of a couple weeks troubleshooting.

 

I swapped my 120gb kingston SSD for a 500gb Samsung QVO for my cache drive. I put the 120 as a second cache pool. I moved all the data from to the array and then onto the new drive. I also upgraded from Beta25 to Beta35. Now, I have a large number of writes on my Cache drive (over 600,000 over a 24hour period). I tried moving my cache to a xfs img and even a directory but no help. The writing continues and it might be worse on the xfs than on the directory. Iotop does not reveal anything of significance really. Is there something i am missing? The only major difference I can see is that the SSDs are formatted differently (the 120 gb shows a MBR: 4KiB-aligned and the 500gb is MBR: 1MiB-aligned). I am aware that there is a massive uptick in this behavior when there is a stream on plex and when there is transcoding but I just assumed this is normal since it has to essentially write the transcoding file somewhere. Any help is appreciated! 

apollo-diagnostics-20201128-2153.zip

I have not been able to identity the issue still. I have disabled a large number of dockers (exception of Plex, Swag, Bitwarden). I am thinking of perhaps moving Plex's app data folder to another drive to see if it really is plex which is the problem.