SPIKE: Investigate Connection Pool Waiters behaviour on create and update

Testing a special version of Juniper (fork from the latest Juniper HF) showed that after switching to a new version of Vertx 4.2.1 and increasing database connection pool to 15 (default is 5) it showed that srs capable of processing consecutive create/update cycles for 5k, 10k, 25k records at chunk_size of 50 (50K requires increasing database connection pool to 20). Screenshots were attached.

5K Create - 13 min/update - 30 min

10 K create - 24 min/Update - 39 min

25 k create - 55 min/25 k update - 2 hr 33 min

Based on today's 12/03/2021 finding import of 5K Update records completed successfully after increasing connection pool size from 5 to 15 SQL connections on new connection pool, that provided with Vertx 4.2.1. So after story to upgrade to new version of RMB need to add story to adjust connection pool size without source code changes for future. New story was created for it: MODSOURCE-420

Per BE discussion: the issue of objects from Connection pool consuming all the memory can be resolved by upgrading to Vertx 4.2.1, which will be done in scope of MODSOURCE-419.

During the testing on PTF env the memory issues are still observed, but caused by a different problem, not related to the Connection pool.  or  could you please create a separate SPIKE for additional investigation and put the details in there? I think this SPIKE can be closed and investigation continued in a different one. Thank you!

CC: