This persistence stores the data in the RAM of the trino-lb server. This implies all queued and running queries will be lost when trino-lb is restarted. Additionally, as multiple trino-lb instances do not share their memory, you can not run multiple trino-lb instances for scalability or availability reasons.
Because of these restrictions the in-memory persistence is only recommended for testing purpose.
The configuration of the in-memory persistence is the most simple, as it does not require any configurations and can be configured as follows:
trinoLb:
persistence:
inMemory: {}