Hi,
I have 2 AsyncFunctions SampleCopyAsyncFunction and SampleSinkAsyncFunction
called with AsyncDataStream.unorderedWait. The 1st
AsyncDataStream.unorderedWait’s
SampleCopyAsyncFunction .asyncInvoke gets called properly but the 2nd
SampleSinkAsyncFunction.asyncInvoke never gets called(though open
OK, I'll close 7470
On Thu, Jul 26, 2018 at 11:25 PM Gary Yao wrote:
> Hi,
>
> Sorry for the late reply. I have seen that you debugged this already and
> created FLINK-9936. Thank you for looking into the issue. I think your
> conclusions are correct. I just wanted to note that there is an even
I recently began using the exactly-once processing semantic with the Kafka
0.11 producer in Flink 1.4.2. It's been working great!
Are Kafka transactions committed when creating a Flink savepoint? How does
this affect the recovery behavior in Flink if, before the completion of the
next checkpoint,
I was trying to find some comparability tables between various versions of
Flink clusters and pipeline jars but haven't run into any so far. Is it so
that the pipeline jars must be build with the same version of the cluster
they'll be running on? Or there are some backward comparability? If it's
al
Hi,
Sorry for the late reply. I have seen that you debugged this already and
created FLINK-9936. Thank you for looking into the issue. I think your
conclusions are correct. I just wanted to note that there is an even older
ticket describing the same problem:
https://issues.apache.org/jira/bro
Hi Chang Liu,
The unbounded nature of the stream keyed or not should not lead to out of
memory.
Flink parallel keyed operator instances have fixed number (parallelism) and
just process some range of keyed elements, in your example it is a subrange of
session ids.
The keyed processed element
