According to PM2 documentation, it claims that the in-built load balancer does not require code modification.Yet, i can't find on which cases this aplies to.In my application, i want a SINGLE worker to handle a SINGLE job. I do not want all workers to do the same job.So, i end up with a custom load-balancer that its role is to assign jobs to the workers (1:1, job:worker) by using the PM2 API (sendDataToProcessId).I have in total 63 clustered workers and 1 fork load-balancer (Thats because i have 64 real threads available).Is this wrong or correct way? Am i missing something?Thanks for your time.
Submitted May 15, 2019 at 08:22PM by pror90
No comments:
Post a Comment