0

We are using akka.net in our project for distributed worker service. We are using Blazor application as UI (actor system with API role) and dot net core worker service multiple instances (same actor system with WLM role). Everything works good however we have issue with service stability. Some of the service nodes goes down intermittently with message - connection is reset by remote peer. In such scenario we are unable to catch the root cause of this disassociation. Please let us know the way to improve the stability of system. Also, is there any way to log the root cause of system disassociation.

Many thanks in advance.

Harshal Sonparote
  • 1
  • 1
  • Could you show some of the code? There's a dozen of reasons why this could happen, most often because remote node has been abruptly stopped working (which is quite common eg. when your actor system was disposed right after starting it because main application thread wasn't blocked after ActorSystem.Start). – Bartosz Sypytkowski Apr 20 '21 at 16:43
  • Thank you Bartosz for responding. I will definately share the logic of actor system creation by tomorrow. However the cluster works good sometime for a day and sometimes for a week. We tried lot of hocon settings like increasing acceptable heartbeat interval e, increasing connection timeout. However nothing is working. Your help will be highly appreciated – Harshal Sonparote Apr 20 '21 at 17:16

0 Answers0