Gyula Fora created FLINK-26572: ---------------------------------- Summary: Re-schedule reconcile more often until job is in ready state Key: FLINK-26572 URL: https://issues.apache.org/jira/browse/FLINK-26572 Project: Flink Issue Type: Sub-task Components: Kubernetes Operator Reporter: Gyula Fora
Currently we only use 2 reconcile reschedule configs. One that is specific to when the port is ready but we need to reschedule one more time (this is 10 seconds by default) and another general reconcile reschedule delay (60 seconds) We should introduce another setting to use when the job is in a deploying or savepointing state to allow for reaching the READY status faster. We could call it: operator.observer.progress-check.interval.sec or operator.observer.operation-check.interval.sec I suggest to use 10 seconds here by default also. -- This message was sent by Atlassian Jira (v8.20.1#820001)