flink-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From Maximilian Michels <...@apache.org>
Subject Re: [DISCUSS] Change semantics of print() to "eager"
Date Tue, 28 Apr 2015 15:27:35 GMT
>
> Why aren't we adding a new method "printLocal()" which is doing the local
> printing?


Simply because it is counter-intuitive that a regular print does not print
on the Client. Introducing a new method would not solve this confusion. I
think it's fine to break the API as long as we provide a meaningful error
message.

It should be ok to use the Accumulator way. We have to force an execution
either way to compute the result to be printed. Printing shouldn't be used
for large output anyways.

On Tue, Apr 28, 2015 at 3:49 PM, Robert Metzger <rmetzger@apache.org> wrote:

> Why aren't we adding a new method "printLocal()" which is doing the local
> printing?
> That would not break any existing code.
>
> How much work would it be to implement this "properly" without piggybacking
> the accumulators?
> I assume we would need to write the data to one or more partitions and
> request the partitions from the client JVM.
> I also assume that we have the code in place now to do it like this.
>
> On Tue, Apr 28, 2015 at 3:45 PM, Till Rohrmann <trohrmann@apache.org>
> wrote:
>
> > +1
> >
> > On Tue, Apr 28, 2015 at 3:19 PM, Maximilian Michels <mxm@apache.org>
> > wrote:
> >
> > > +1 Very nice addition.
> > >
> > > On Tue, Apr 28, 2015 at 2:12 PM, Stephan Ewen <sewen@apache.org>
> wrote:
> > >
> > > > Sounds good, Max, let's to this in one fix.
> > > >
> > > > We can maintain a counter in the ExecutionEnvironment that tracks how
> > > many
> > > > executions have happened.
> > > > In case of no prior execution, simply warn that no sinks are defined.
> > > > In case a prior execution happened, point out that nothing new is
> > pending
> > > > execution.
> > > >
> > > > On Tue, Apr 28, 2015 at 2:04 PM, Maximilian Michels <mxm@apache.org>
> > > > wrote:
> > > >
> > > > > I agree, print should print on the client. However, let's introduce
> > > some
> > > > > big hint in the error message in case of a second execute() that
> this
> > > > error
> > > > > may arise from a previous execution.
> > > > >
> > > > > Instead of "No sinks defined", let's print "The Flink job didn't
> > > contain
> > > > > any sinks. This may be because the sinks were already executed. If
> > you
> > > > > executed the print() method on a DataSet before, the job would have
> > > > already
> > > > > been executed. In this case, remove the call of execute() until you
> > > have
> > > > > defined further sinks".
> > > > >
> > > > > On Tue, Apr 28, 2015 at 1:24 PM, Fabian Hueske <fhueske@gmail.com>
> > > > wrote:
> > > > >
> > > > > > +1 for the breaking change
> > > > > >
> > > > > > 2015-04-28 13:18 GMT+02:00 Ufuk Celebi <uce@apache.org>:
> > > > > >
> > > > > > >
> > > > > > > On 28 Apr 2015, at 12:31, Stephan Ewen <sewen@apache.org>
> wrote:
> > > > > > >
> > > > > > > > +1 for the breaking change
> > > > > > > >
> > > > > > > > Let's not to this any more than necessary, bu this
is a good
> > > > case...
> > > > > > >
> > > > > > > +1
> > > > > > >
> > > > > >
> > > > >
> > > >
> > >
> >
>

Mime
  • Unnamed multipart/alternative (inline, None, 0 bytes)
View raw message