Ensure operators catch the shutdown signal #803
Description
At least secret-operator (in controller mode) doesn't catch CTRL+C (SIGINT), so when k8s is trying to shutdown the pod, it times out and gets force killed (I assume SIGKILL, at which point the code has no choice in the matter).
Tasks
Tip
This needs a little bit of thought around signalling shutdown to nested futures (eg: kube controller) so that the shutdown is graceful. IMO, the signal should be caught in fn main() and then use a channel to signal to nested things when a shutdown is requested.
- Go through all operators and ensure they can be shutdown via common unix signals.
- feat: Add graceful shutdown airflow-operator#741
- feat: Add graceful shutdown druid-operator#788
- feat: Add graceful shutdown hbase-operator#730
- feat: Add graceful shutdown hdfs-operator#747
- feat: Add graceful shutdown hive-operator#670
- feat: Add graceful shutdown kafka-operator#934
- feat: Add graceful shutdown nifi-operator#894
- feat: Add graceful shutdown opensearch-operator#110
- feat: Add graceful shutdown spark-k8s-operator#651
- feat: Add graceful shutdown superset-operator#699
- feat: Add graceful shutdown trino-operator#843
- feat: Add graceful shutdown zookeeper-operator#1004
- feat: Add graceful shutdown commons-operator#396
- feat: Add graceful shutdown listener-operator#366
- feat: Add graceful shutdown opa-operator#804
- feat: Add graceful shutdown secret-operator#674