RE: My Looper job iterations are staying in QUEUE state after first 10 iterations are completed
I have 180 iterations returned by Looper and I have setup the parallel limit as 10 in looper start. After starting the looper the first 10 iterations completed successfully and after that all the jobs are staying in QUEUE change.
Iterations in QUEUE:
Below are the recommended configurations at the controller and Admin level to overcome this IN QUEUE issue when jobs are running in parallel.
- In the postgresql.conf file present at location [diyotta/controller-repo/direpdata/] Change the max_connections value to 1000 (max_connections =1000).
- In the diserver.config file present at location [diyotta/controller/conf/] change the hibernate.c3p0.max_size value to 850 (hibernate.c3p0.max_size = 850)
- In the admin module move to the property [Admin à Configurations à Diyotta Constants Settings ] change the ops.OPS_NO_OF_THREADS value to 800 (ops.OPS_NO_OF_THREADS = 800).
- When we are trying to run the parallel instances using Looper make sure we are running 10 parallel instances at once.
Note: After making the above changes, restart the controller and then run the jobs.
Ideally consider this max_connections > hibernate.c3p0.max_size > OPS_NO_OF_THREADS
OPS_NO_OF_THREADS should be less then hibernate.c3p0.max_size. If you are making max_connection to 1000 then change max_size to 850 or 900.. keep the difference of at least 150-100 connections