Hello , I am deploying flink job using flink-kuberentes-operator in application mode. My requirement is to achieve quicker recovery during any TM failures or exceptions. I want to have have stand by slots (TM pods) so that if one TM is lost, job can find a free slot immediately. Restart strategy is 5m - 20 attempts. If we dont have backup//extra slots, then I see job waiting for a longer time slots be to provisioned. Using 1slot per TM. How can I have configure FlinkDeployment to have extra slots. Need some inputs on this.

you can override slot using flinkConfiguration section in FlinkDeployment spec

Flink Job manager always spins us TM pods based on number of slots and the maximum operator parallelism of the job. I can adjust slot number to achieve this config. But 1slot 1TM model is working fine for me and dont want to change. eg: If job parallelism is 24 then I can set slots as 5 then i get 5TMs which will give me 1 back up