Event persist error as active persistent actor count increase


I’m trying to profile how many active persistent actors in a node/pod with 1 GiB memory can be supported. I find that I reach around 50k active persistent actors then I see db persist errors like this, “akka.actor.default-dispatcher-7] - Supervisor StopSupervisor saw failure: Failed to persist event type”. I am using Minikube with 8GB allocated on my local machine. Isn’t 50k a bit too small for 1GiB of pod/node memory? I’ve tried adding delays between sending event triggering messages to the actors. I’ve tried decreasing the parallel request from 100 to 20. To no avail. I encountered some logs indicating GC is taking a bit too long. Is there a config I can change that can improve the performance? Like maybe decrease the thread pool of the db?