Github user aozturk commented on a diff in the pull request:
https://github.com/apache/flink/pull/2016#discussion_r64756634
--- Diff:
flink-streaming-connectors/flink-connector-kinesis/src/main/java/org/apache/flink/streaming/connectors/kinesis/FlinkKinesisProducer.java
---
@@ -160,12 +168,13 @@ public void
setCustomPartitioner(KinesisPartitioner<OUT> partitioner) {
public void open(Configuration parameters) throws Exception {
super.open(parameters);
- KinesisProducerConfiguration config = new
KinesisProducerConfiguration();
- config.setRegion(this.region);
- config.setCredentialsProvider(new StaticCredentialsProvider(new
BasicAWSCredentials(this.accessKey, this.secretKey)));
+ KinesisProducerConfiguration producerConfig = new
KinesisProducerConfiguration();
+
+
producerConfig.setRegion(configProps.getProperty(KinesisConfigConstants.CONFIG_AWS_REGION));
+
producerConfig.setCredentialsProvider(AWSUtil.getCredentialsProvider(configProps));
//config.setCollectionMaxCount(1);
//config.setAggregationMaxCount(1);
--- End diff --
Sorry again for my late response. Meanwhile I have thought about this and I
have come up with a question. I had noticed a remark about not using the KCL
library, but is there any special reason why we are not using KPL? Anything we
intend to do will probably replicate it including retry mechanism on failures,
multi-threading and asynchronous writes.
We also need to take into consideration that our producer solution should
be cost effective by using the opportunities for batching, which can be
customized by the user if low latency is preferred.
---
If your project is set up for it, you can reply to this email and have your
reply appear on GitHub as well. If your project does not have this feature
enabled and wishes so, or if the feature is enabled but not working, please
contact infrastructure at infrastruct...@apache.org or file a JIRA ticket
with INFRA.
---
