THIS IS A TEST INSTANCE. ALL YOUR CHANGES WILL BE LOST!!!!
...
- Support autoscaling Flink jobs deployed on Kubernetes via the Flink Kubernetes operator
- Provide a default Flink job metric collector via the Flink metrics API
- Provide a robust default scaling algorithm
- Ensure scaling yields effective usage of assigned task slots, i.e. prevents backpressure but does not overprovision
- Ramp up in case of any backlog to ensure it gets processed in a timely manner
- Minimize the number of scaling decisions to prevent costly rescale operation
- Allow plugging in a custom scaling algorithm and metric collection code
Non-Goals (for now)
- Vertical scaling via adjusting container resources or changing the number of task slots per TM
- Autoscaling in standalone Flink session cluster deployment mode
- Integrate with the k8s scheduler to reserve required resources for rescaling
...