Hi Daniel,
The answer to you original question is you can just keyBy[1] by e.g. the
machineId and then computations on KeyedStream are applied independently
for each key.
Best,
Dawid
[1]
https://ci.apache.org/projects/flink/flink-docs-release-1.7/dev/stream/operators/#datastream-transformations
I don't get what happened here. Did Selvaraj just hijack this question? Or
what is going on?
Am Di., 29. Jan. 2019 um 17:01 Uhr schrieb Selvaraj chennappan <
selvarajchennap...@gmail.com>:
> I think there is misunderstanding . I want to compare raw json and
> transformed record .
> Hence I need t
I think there is misunderstanding . I want to compare raw json and
transformed record .
Hence I need two consumer and merge the stream for comparison.
I have pipeline defined . pipeline does source(kafka)
,transformation,dedup and persisting to DB .
[image: image.png]
Before reaching to DB task l
Hi Selvaraj
In your pojo add data member as status or something like that,now set it
error in case it is invaild .pass the output of flatmap
to split opertor there you can split the stream
On Tue, Jan 29, 2019 at 6:39 PM Selvaraj chennappan <
selvarajchennap...@gmail.com> wrote:
> UseCase:- We h
UseCase:- We have kafka consumer to read messages(json ) then it applies to
flatmap for transformation based on the rules ( rules are complex ) and
convert it to pojo .
We want to verify the record(pojo) is valid by checking field by field of
that record .if record is invalid due to transformation
Im not sure if i got your question correctly, can you elaborate more on
your use case
