Import of 5 concurrent jobs become stuck

This issue is observed in a Honeysuckle HF3 FOLIO environment.

Creating ticket with context to check the same scenario does not exist in Iris environment

Specifically - 5 parallel Data Import job runs were submitted at that time, jobs progressed for about 30 mins or so and have been stuck ever since.

Observe errors in mod-pubsub audit_message table indicating some events were not successfully delivered (most likely due to issues with inventory at the time)Error delivering DI_SRS_MARC_BIB_INSTANCE_HRID_SET event with id 'ad1a0390-7926-4533-b9a7-044f80d170c9' to /inventory/handlers/instances, response status code is 503, Service Unavailable

As well as messages in the log

Error delivering DI_SRS_MARC_BIB_INSTANCE_HRID_SET event with id '22d302a2-36c0-4ac0-9618-037ee6a33806' to /inventory/handlers/instances, response status code is 502, Bad Gateway

And message from mod-source-record-storage
20:15:00.081 [vert.x-eventloop-thread-0] ERROR tpClientResponseImpl [10600064eqId] io.vertx.core.VertxException: Connection was closed

Spikes in CPU and Memory were observed for both mod-inventory and mod-pubsub

After Restarting modules – attempts to run concurrent imports continued to get stuck