#[0]
#[0]
#[0]
#[0]
#[0]
#[0] ah no clue and I refuse to debug gay features
#[0]
#[0]
#[0]
#[0]
#[0] :licc:
#[0]
#[0]
#[0] :licc:
#[0]
#[0]
#[0]
#[0] milk and honey
Didn't you read the Bible?
#[0] #[1] #[2] #[3] #[4] #[5] I did
#[0] #[1] #[2] #[3] biggest thing is match the block size of the SSD, be sure to enable 4kn blocks on the SSD, set the record size to 16k. Then there's a good bit of tuning to do in postgres. Also, don't fall for the zstd compression meme. Too much latency for something like pleroma. Use lz4.
Here's a good resource:
https://vadosware.io/post/everything-ive-seen-on-optimizing-postgres-on-zfs-on-linux/
In a VM you lose a lot of the benefits and I don't see much of a point in using ZFS. Now, using ZFS for the underlying VM storage makes sense.
ZFS really is meant to be used on bare metal. I think this is where a majority of your problems are coming from.