Hey team 👋 When using the ElasticSearch sink, is there a way to dynamically change

bulkFlushMaxActions

parameter? Say that due to ES struggling to index a bulk of 1000 actions and slow response, I'm thinking to have Flink adjust the bulk size based on the latency from ES, so if it takes 5s to index 1k actions, next bulk should be 500 actions and so on... Is this doable in Flink?

Following the docs it is not possible ATM. But we could create some sort of back-off strategy based on the response delay. IIRC, the config is only meant to be a “hard limit”, so no extra computations would be required.