Hello,
I have postgres as source and mysql as sink. The user authenticating with
mysql does _not_ have DELETE privileges.
In some cases, flink throws error because it is trying to _delete_ records
and the user does not have privilege. In most cases (of the same job),
upsert is working as expected
Hey wang!
Perhaps this is what you want:
https://nightlies.apache.org/flink/flink-docs-release-1.14/docs/connectors/table/kafka/#key-format
&
https://nightlies.apache.org/flink/flink-docs-release-1.14/docs/connectors/table/kafka/#key-fields
?
Note that the fields *have* to be one of the "top" lev
> was fixed recently and so it's still not released.
>
> Regards,
> Dian
>
> On Thu, May 12, 2022 at 7:26 PM Dhavan Vaidya
> wrote:
>
>> Hey Dian,
>>
>> Though my HTTP call's response is indeed JSON, I needed to serialize data
>> into A
Hi Ian,
You need to create two streams, and use
https://nightlies.apache.org/flink/flink-docs-master/api/python/pyflink.datastream.html#pyflink.datastream.DataStream.connect
to "join" them.
On Fri, 13 May 2022 at 14:06, lan tran wrote:
> Hi team, I have the use case is that I want to join two d
o me that having `Row`
> inside `Row` is causing some issues.
> It supports `Row` inside `Row`. Could you share an example which could
> reproduce this issue?
>
> Regards,
> Dian
>
>
> On Tue, May 10, 2022 at 9:09 PM Dhavan Vaidya
> wrote:
>
>> Hel
Hello,
I am consuming Kafka messages with Table API connector. I cannot use
DataStream API because it does not support Confluent Avro.
After consuming the messages, I am converting to DataStream API and using
ProcessFunction. The ProcessFunction makes async http calls and emits
results with a com
cio operator supported in
> Flink, e.g. you are waiting the results of one element before processing
> the next element and so it's actually synchronous from this point of view.
>
> Regards,
> Dian
>
> On Thu, May 5, 2022 at 9:52 PM Dhavan Vaidya
> wrote:
>
>&g
rator is still not supported in PyFlink.
>
> Regards,
> Dian
>
> On Tue, May 3, 2022 at 3:48 PM Dhavan Vaidya
> wrote:
>
>> Hey Francis!
>>
>> Thanks for the insights! I am thinking of using Java / Scala for this
>> scenario given your input. Introducing a
gt; might be to use statefun for the enrichment stages. We've also changed our
> model for enrichment, we're pushing the enrichment data into the pipeline
> instead of pulling it, but this won't work in a lot of situations.
>
> Hope that gives you some ideas.
>
> On
Hello!
I want to make HTTP(S) calls to enrich data streams. The HTTP services are
running on our VPC, so the delay is limited, but sometimes these services
end up calling third party APIs, and latencies become high.
>From documentation (
https://nightlies.apache.org/flink/flink-docs-release-1.14/
10 matches
Mail list logo
