Hi Camunda Team,
I am trying to fine tune my camunda 8.5 env to handle max load.
I have 11 broker(4gb and 2 cpu each), 22 partitions
and for Zeebe Client have 3 pod count (4 gb memory for each) which can go upto 6
I my process implementation is complex and i have job workers(aroung 20-25 workers) each having api calls and DB call so it takes time to process…
zeebe.client.worker.maxJobActive is 32
job.pollInterval is 0.1 sec
and worker.threads was initially 5 but i increased it to 15, still i dont see much improvement.
Is there anything wrong I am doing?
what could be the possible values for these variables so that it can process more number of instances at time.
with 5k instances it is taking about an hour to complete, we need to reduce it to 10min. And zeebe client is only using average 2gb of memory only