davidradl commented on code in PR #27163:
URL: https://github.com/apache/flink/pull/27163#discussion_r2472641981
##########
docs/content/docs/connectors/models/openai.md:
##########
@@ -157,6 +157,57 @@ FROM ML_PREDICT(
</ul>
</td>
</tr>
+ <tr>
+ <td>
+ <h5>error-handling-strategy</h5>
+ </td>
+ <td>optional</td>
+ <td style="word-wrap: break-word;">retry</td>
+ <td>Enum</td>
+ <td>Strategy for handling errors during model requests. Supported
strategies:
+ <ul>
+ <li><code>retry</code>: Retry sending the request. The
retrying behavior is limited by retry-num, retry-fallback-strategy,
retry-backoff-strategy and retry-backoff-base-interval.</li>
+ <li><code>failover</code>: Throw exceptions and fail the
Flink job.</li>
+ <li><code>ignore</code>: Ignore the input that caused the
error and continue. The error itself would be recorded in log.</li>
Review Comment:
For the [HTTP
connector](https://github.com/apache/flink-connector-http/blob/main/docs/content/docs/connectors/table/http.md)
. We have introduced metadata columns to surface the error details and a flag
to continueOnError. Can we have a strategy here that would surface the error
and carry on. This allows the stream to "handle" the error with dead letter
queues etc.
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
To unsubscribe, e-mail: [email protected]
For queries about this service, please contact Infrastructure at:
[email protected]
