3280895 - Retry of JMS messages is delayed

If a message starts from a queue and processing fails, the message is stored in an internal error queue. Each processing queue has its own error queue. A periodic job checks these messages according to the configured retry interval and moves the message back to the processing queue. This job runs once a minute.

• In the Neo environment, the job runs on each tenant management node.
• In the Cloud Foundry environment, the job runs on each worker node.

The job uses the JMS application programming interface (API) to read messages from each error queue. This API requires the setting of a timeout, the maximum waiting time for a message. That means, if an error queue is empty, the API call takes exactly this maximum waiting time. The default value for the timeout is 5 seconds. Starting from that value, the job can calculate that already for more than 12 queues the job execution time is larger than the scheduled job period of one minute.

SAP BTP Neo and Cloud Foundry environments

KBA , LOD-HCI-PI-CON-SOAP , SOAP Adapter , Problem