Ahmed Hamdy created FLINK-40695:
-----------------------------------
Summary: Transient issues during ExecutionGraph creation lead to
terminal job Failure
Key: FLINK-40695
URL: https://issues.apache.org/jira/browse/FLINK-40695
Project: Flink
Issue Type: Bug
Components: Runtime / Coordination
Affects Versions: 2.2.1
Reporter: Ahmed Hamdy
During job submission or rescaling the AdaptiveScheduler moves to
{{CreatingExecutionGraph}} state which handles failures by moving to the
terminal {{Failed}} state. Failures in execution graph creation can be
transient or retryable like network issues preparing checkpoint directories.
h2. Acceptance Criteria
Introduce a retry mechanism for ExecutionGraph creation
--
This message was sent by Atlassian Jira
(v8.20.10#820010)
