[
https://issues.apache.org/jira/browse/FLINK-35989?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
]
Hong Liang Teoh updated FLINK-35989:
------------------------------------
Fix Version/s: aws-connector-5.0.0
> Document and log error on partially failed requests for AWS sinks
> -----------------------------------------------------------------
>
> Key: FLINK-35989
> URL: https://issues.apache.org/jira/browse/FLINK-35989
> Project: Flink
> Issue Type: Improvement
> Components: Connectors / DynamoDB, Connectors / Firehose, Connectors
> / Kinesis
> Affects Versions: aws-connector-4.3.0
> Reporter: Hong Liang Teoh
> Priority: Major
> Fix For: aws-connector-4.4.0, aws-connector-5.0.0
>
>
> h2. Problem
> Currently, AWS sinks send batch write requests to endpoints.
> On partial failure within the batch (only some records fail), we perform
> retries or fail the job, but we don't log the error returned.
> {code:java}
> private void handlePartiallyFailedRequest(
> PutRecordsResponse response,
> List<PutRecordsRequestEntry> requestEntries,
> Consumer<List<PutRecordsRequestEntry>> requestResult) {
> LOG.warn(
> "KDS Sink failed to write and will retry {} entries to KDS",
> response.failedRecordCount());
> numRecordsOutErrorsCounter.inc(response.failedRecordCount());
> if (failOnError) {
> getFatalExceptionCons()
> .accept(new
> KinesisStreamsException.KinesisStreamsFailFastException());
> return;
> }
> List<PutRecordsRequestEntry> failedRequestEntries =
> new ArrayList<>(response.failedRecordCount());
> List<PutRecordsResultEntry> records = response.records();
> for (int i = 0; i < records.size(); i++) {
> if (records.get(i).errorCode() != null) {
> failedRequestEntries.add(requestEntries.get(i));
> }
> }
> requestResult.accept(failedRequestEntries);
> }{code}
> Code link:
> [https://github.com/apache/flink-connector-aws/blob/main/flink-connector-aws/flink-connector-aws-kinesis-streams/src/main/java/org/apache/flink/connector/kinesis/sink/KinesisStreamsSinkWriter.java#L254-L274]
> h2. What we want to do
> We should log the reason for failure to assist user debugging.
--
This message was sent by Atlassian Jira
(v8.20.10#820010)
