cross-posted from: https://swg-empire.de/post/4511580
In my relentless pursuit of trying to coax more performance out of my Lemmy instance I read that PostgreSQL heavily relies on the OSs disk cache for read performance. I’ve got 16 GB of RAM and two hdds in RAID 1. I’ve PostgreSQL configured to use 12 GB of RAM and I’ve zram swap set up with 8 GB.
But according to htop PostgreSQL ia using only about 4 GB. My swap gets hardly touched. And read performance is awful. Opening my profile regularly times out. Only when it’s worked ones does it load quickly until I don’t touch it again for half an hour or so.
Now, my theory is that the zram actually takes available RAM away from the disk cache, thus slowing the whole system down. My googling couldn’t bring me the answer because it only showed me how to set up zram in the first place.
Does anyone know if my theory is correct?
You’re probably correct, although I have no experience with zram so can’t be sure. But you’re absolutely right that PostgreSQL depends heavily on the OS disk cache for optimal performance. Lowering the PostgreSQL setting like Blaster M suggests won’t improve performance much, since all that setting does is tell PostgreSQL’s algorithms how much memory is likely to be allocated to the OS disk cache. Of course it’s best if it’s accurate, so you’re best off seeing how much memory is actually allocated to disk cache under heavy use before setting it, but it shouldn’t massively reduce performance if you don’t get it right.
Lower the postgre to 8GB and see what happens? Also, hard drives, ssds, or nvme ssds? Recent info suggests it is possible memcaching is actually slower than direct access to nvme ssd
