Hyukjin Kwon created SPARK-61149:
------------------------------------
Summary: Retry by the policy past the bound of RetryExceptions,
take SESSION_NOT_FOUND and a response already received in reattach, and bound
the config read of an error, as PySpark does
Key: SPARK-61149
URL: https://issues.apache.org/jira/browse/SPARK-61149
Project: Spark
Issue Type: Bug
Components: Connect
Affects Versions: connect-rust-4.3.0
Reporter: Hyukjin Kwon
The Rust client's policy retries stop once the one-hour bound of
RetryExceptions has passed, which PySpark applies to RetryExceptions only, so a
transient UNAVAILABLE on a long reattach fails with retries left.
An unknown session is not handled like an unknown operation, and a reattach
that reads an unknown operation after a response raises the server error, not
RESPONSE_ALREADY_RECEIVED as PySpark does.
The config read that shows the stack trace of an error has no deadline, so a
half-open connection hangs the call that should raise the error.
--
This message was sent by Atlassian Jira
(v10.3.26#10030026)
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
