About 2gb and it should be distributed
בתאריך יום א׳, 15 במאי 2022, 19:50, מאת Reuven Lax :
> How large is this state? Is it distributed?
>
> On Sun, May 15, 2022 at 8:12 AM Sigalit Eliazov
> wrote:
>
>> Thanks for your response.
>> The use case is 2 pipelines:
>> A. receives information from k
Thanks very much for your attention to this problem!
Deepak
On Mon, May 16, 2022 at 1:42 PM Robert Bradshaw wrote:
>
> I've filed https://issues.apache.org/jira/browse/BEAM-14476 to track this.
>
> On Mon, May 16, 2022 at 1:30 PM Deepak Nagaraj
> wrote:
> >
> > I just tried DirectRunner -- I di
I just tried DirectRunner -- I did not see any problem with it. It
does happen with Flink.
Deepak
On Mon, May 16, 2022 at 12:57 PM Robert Bradshaw wrote:
>
> Is this just on Flink, or does this happen on other runners too?
>
> On Mon, May 16, 2022 at 12:39 PM Deepak Nagaraj
> wrote:
> >
> > Hi
Hi Robert,
On Mon, May 16, 2022 at 12:33 PM Robert Bradshaw wrote:
>
> On Mon, May 16, 2022 at 12:01 PM Deepak Nagaraj
> wrote:
> >
>
> I can imagine contention for an I/O lock, but I'm not sure how that
> would lead to a deadlock. But hopefully knowing that print() is
> involved should allow a
Hi Robert,
Thanks, a coworker and I have narrowed this down to a print(msg) statement
in one of my beam.Map() functions. If we remove the print(), the pipeline
does not stall. The print() itself is necessary but not sufficient: the
problem occurs only when there are many large messages coming into
Hi there,
I’m trying out a streaming pipeline which reads events from a pubsub topic and
I want to pass a side input to the stage that processes this pubsub event,
updating the side input value every 5 minutes using an impulse that kicks off a
time.Ticker which emits every 5 minutes.
The strea