@matty @matty @matty @alex definitely something not right with the workers if thats the case. it took me like a week of messing with queues to get the delicate balance we needed. i think i ended up on 30 for notifications and 100/100 for federator incoming/outgoing. i wouldnt suggest that at all for you, ours is that high due to traffic + postgresql being on its own epyc server