= IntStream.rangeClosed(1,
> inputSize).boxed().collect(Collectors.toList());
>
> // 1st half of pipeline
> //DataStream> pipeA = env.fromCollection(listOfNumbers)
> DataStream> pipeA = env.addSource(new
> StreamTest.DripFeed(inputSize))
> .windowA
___
From: Biao Liu
Sent: Tuesday 17 December 2019 21:50
To: John Morrow
Cc: user
Subject: Re: MiniCluster with ProcessingTimeTrigger
Hi John,
The root cause is the collection source exits too fast. The window would also
exit without being triggered.
You could verify that by waiting
Hi John,
The root cause is the collection source exits too fast. The window would
also exit without being triggered.
You could verify that by waiting a second before releasing the window. For
example, insert a map operator between source and window operator. Blocking
a second or more in the
Hi All,
I'm trying to test a pipeline that consists of two Flink tasks with a
MiniCluster. The 1st task has a WindowAll operator which groups items into
batches every second, and the 2nd task does an async operation with each batch
and flatMaps the result.
I've whittled it down to the bare
