[GitHub] flink issue #5563: [FLINK-8543] Don't call super.close() in AvroKeyValueSink...

2018-02-26 Thread aljoscha 
Github user aljoscha commented on the issue:

https://github.com/apache/flink/pull/5563
  
merged


---

[GitHub] flink issue #5563: [FLINK-8543] Don't call super.close() in AvroKeyValueSink...

2018-02-26 Thread aljoscha 
Github user aljoscha commented on the issue:

https://github.com/apache/flink/pull/5563
  
@zentol I pushed an update.


---

[GitHub] flink issue #5563: [FLINK-8543] Don't call super.close() in AvroKeyValueSink...

2018-02-26 Thread aljoscha 
Github user aljoscha commented on the issue:

https://github.com/apache/flink/pull/5563
  
@IgorBerman Thanks! I'm expecting no problems, just wanted to see if you 
maybe had an opinion.


---

[GitHub] flink issue #5563: [FLINK-8543] Don't call super.close() in AvroKeyValueSink...

2018-02-26 Thread IgorBerman 
Github user IgorBerman commented on the issue:

https://github.com/apache/flink/pull/5563
  
@aljoscha, sorry for late response, I needed to refresh this piece of code. 
Overall, I think it's ok. I mean at the end I've used same scheme of 
implementing SinkWriter as other examples, so if you refactoring something in 
hierarchy then it should have same handling.
From avro file perspective(internal writer) you are closing it, which would 
be sufficient. What problems do you expect here?


---

[GitHub] flink issue #5563: [FLINK-8543] Don't call super.close() in AvroKeyValueSink...

2018-02-22 Thread aljoscha 
Github user aljoscha commented on the issue:

https://github.com/apache/flink/pull/5563
  
@IgorBerman Do you have an idea whether this change could cause problems? 
As the original implementer.



---