This already sounds awfully complicated. Is there no other way to
implement the delta windows?
On Wed, Jun 3, 2015 at 7:52 AM, Gyula Fóra wrote:
> Hi Ufuk,
>
> In the concrete use case I have in mind I only want to send events to
> another subtask of the same task vertex.
>
> Specifically: if we
Hi Ufuk,
In the concrete use case I have in mind I only want to send events to
another subtask of the same task vertex.
Specifically: if we want to do distributed delta based windows we need to
send after every trigger the element that has triggered the current window.
So practically I want to br
Hello Ufuk, Robert,
@Ufuk if you can create a ticket and assign it to me that would be very
helpful
@Robert I can definitely update those two documents. Just now I was
checking and looks like we can control this using checkstyle
http://checkstyle.sourceforge.net/config_imports.html#IllegalImport
Márton Balassi created FLINK-2139:
-
Summary: Test Streaming Outputformats
Key: FLINK-2139
URL: https://issues.apache.org/jira/browse/FLINK-2139
Project: Flink
Issue Type: Test
Compo
Márton Balassi created FLINK-2138:
-
Summary: PartitionCustom for streaming
Key: FLINK-2138
URL: https://issues.apache.org/jira/browse/FLINK-2138
Project: Flink
Issue Type: New Feature
Márton Balassi created FLINK-2137:
-
Summary: Expose partitionByHash for WindowedDataStream
Key: FLINK-2137
URL: https://issues.apache.org/jira/browse/FLINK-2137
Project: Flink
Issue Type: New
Márton Balassi created FLINK-2136:
-
Summary: Test the streaming scala API
Key: FLINK-2136
URL: https://issues.apache.org/jira/browse/FLINK-2136
Project: Flink
Issue Type: Test
Compo
If it helps you with your task, then you can add it. The best thing is
probably to implement it similarly to the mapWithBcVariable.
Cheers,
Till
On Tue, Jun 2, 2015 at 7:25 PM, Sachin Goel
wrote:
> Should I go ahead and add this method then? The mapWithBcSet I mean.
>
> Regards
> Sachin Goel
On 02 Jun 2015, at 22:45, Gyula Fóra wrote:
> I am wondering, what is the suggested way to send some events directly to
> another parallel instance in a flink job? For example from one mapper to
> another mapper (of the same operator).
>
> Do we have any internal support for this? The first thin
Hi,
I am wondering, what is the suggested way to send some events directly to
another parallel instance in a flink job? For example from one mapper to
another mapper (of the same operator).
Do we have any internal support for this? The first thing that we thought
of is iterations but that is clea
Thanks Robert and Stephan.
I executed the program in maven environment. It works fine.
However, in Plugin environment, though i have included all the dependencies
and dependent jars, it says classnotfoundexception while executing the env.
--
View this message in context:
http://apache-flink
Adding an entry here: http://flink.apache.org/coding-guidelines.html is
certainly good, yes.
You can contribute to the website here: https://github.com/apache/flink-web
We enforce coding guidelines using the maven checkstyle plugin. Maybe there
is a way of forbidding certain imports
On Tue, Jun 2
On 02 Jun 2015, at 21:18, Lokesh Rajaram wrote:
> Hello Robert,
>
> I worked on that issue, if it's ok I can take this task.
>
> Btw, how is anything enforced in Flink? Do I have to update how to contribute
> guide or any thing else need to be done?
The how to contribute guide is a good star
Hello Robert,
I worked on that issue, if it's ok I can take this task.
Btw, how is anything enforced in Flink? Do I have to update how to contribute
guide or any thing else need to be done?
Sent from my iPhone
> On Jun 2, 2015, at 12:11 PM, Robert Metzger wrote:
>
> We have now replaced all
We have now replaced all commons validate calls with guava preconditions
but its not written down anywhere or enforced by anything.
Who would like to take care of that?
On Thu, Mar 26, 2015 at 11:03 AM, Robert Metzger
wrote:
> I didn't know that there was already an issue for this. I closed
> F
On 02 Jun 2015, at 20:12, Alexander Alexandrov
wrote:
> During an offline chat some time ago Stephan Ewen mentioned that there is
> an ongoing effort for a dynamic memory allocation in some feature branch
> lying around. Can you point me to that, as I would like to look at the
> code? Thanks.
During an offline chat some time ago Stephan Ewen mentioned that there is
an ongoing effort for a dynamic memory allocation in some feature branch
lying around. Can you point me to that, as I would like to look at the
code? Thanks.
I'm looking into the last one already.
On Tuesday, June 2, 2015, Sachin Goel wrote:
> Ah. This is a known issue then.
>
> Regards
> Sachin Goel
>
> On Tue, Jun 2, 2015 at 9:14 PM, Aljoscha Krettek >
> wrote:
>
> > The last one is actually SuccessAfterNetworkBuffersFailureITCase which
> > uses C
Should I go ahead and add this method then? The mapWithBcSet I mean.
Regards
Sachin Goel
On Tue, Jun 2, 2015 at 10:43 PM, Till Rohrmann
wrote:
> Yes you’re right Sachin. The mapWithBcVariable is only syntactic sugar if
> you have a broadcast DataSet which contains only one element. If you have
Yes you’re right Sachin. The mapWithBcVariable is only syntactic sugar if
you have a broadcast DataSet which contains only one element. If you have
multiple elements in your DataSet then you can’t use this method. But we
can define another method mapWithBcSet which takes a function f: (element:
T,
Further, I think we should return just
broadcastVariable = getRuntimeContext.
getBroadcastVariable[B]("broadcastVariable")
in BroadcastSingleElementMapper
User may wish to have a list broadcasted, and not just want to access the
first element. For example, this would make sense in the kmeans algori
Hi,
the difference is that the IT (Integration Test) suites are only run
when "mvn verify" or "mvn install" is executed. The other suites (or
tests) are ran when "mvn test" is executed. Typically the non-IT tests
are more lightweight while the IT tests can do heavy stuff that takes
a bit longer.
W
Hi
What's the convention with using an "ITSuite" in the name instead of just
"Suite"?
Regards
Sachin Goel
Robert Metzger created FLINK-2135:
-
Summary: Java plan translation fails with ClassCastException
(probably in first())
Key: FLINK-2135
URL: https://issues.apache.org/jira/browse/FLINK-2135
Project: Fl
Ah. This is a known issue then.
Regards
Sachin Goel
On Tue, Jun 2, 2015 at 9:14 PM, Aljoscha Krettek
wrote:
> The last one is actually SuccessAfterNetworkBuffersFailureITCase which
> uses ConnectedComponents.
>
> On Tue, Jun 2, 2015 at 5:42 PM, Robert Metzger
> wrote:
> > Hi,
> > I think there
The last one is actually SuccessAfterNetworkBuffersFailureITCase which
uses ConnectedComponents.
On Tue, Jun 2, 2015 at 5:42 PM, Robert Metzger wrote:
> Hi,
> I think there are currently at least three tests randomly failing on Travis:
> - the yarn failure you've just reported
> - the KafkaITCase
Hi,
I think there are currently at least three tests randomly failing on Travis:
- the yarn failure you've just reported
- the KafkaITCase
- a ConnectedComponent test (Ufuk posted something about this 2 hrs ago)
On Tue, Jun 2, 2015 at 5:41 PM, Sachin Goel
wrote:
> I recently merged the master br
I recently merged the master branch to something I've been working on.
However, the travis build is failing with this error:
I have made no changes to the existing files, only added some new files and
the tests for those actually pass properly.
The build is here: https://travis-ci.org/apache/flink/
Hi Till
This works only when there is only one variable to be broadcasted, doesn't
it? What about the case when we need to broadcast two? Is it advisable to
create a BroadcastDoubleElementMapper class or perhaps we could just send a
tuple of all the variables? Perhaps that is a better idea.
Regard
By the way, we also should rename the corresponding Streaming API
method accordingly.
On Tue, Jun 2, 2015 at 3:24 PM, Maximilian Michels wrote:
> +1 for printOnTaskManager(prefix)
>
> On Tue, Jun 2, 2015 at 1:54 PM, Kostas Tzoumas wrote:
>
>> +1 for printOnTaskManager(prefix)
>>
>> On Tue, Jun 2
+1 for printOnTaskManager(prefix)
On Tue, Jun 2, 2015 at 1:54 PM, Kostas Tzoumas wrote:
> +1 for printOnTaskManager(prefix)
>
> On Tue, Jun 2, 2015 at 1:35 PM, Till Rohrmann
> wrote:
>
> > +1 for printOnTaskManager(prefix)
> >
> > On Tue, Jun 2, 2015 at 12:08 PM, Fabian Hueske
> wrote:
> >
> >
Ufuk Celebi created FLINK-2134:
--
Summary: Deadlock in SuccessAfterNetworkBuffersFailureITCase
Key: FLINK-2134
URL: https://issues.apache.org/jira/browse/FLINK-2134
Project: Flink
Issue Type: Bug
Aljoscha Krettek created FLINK-2133:
---
Summary: Possible deadlock in ExecutionGraph
Key: FLINK-2133
URL: https://issues.apache.org/jira/browse/FLINK-2133
Project: Flink
Issue Type: Bug
+1 for printOnTaskManager(prefix)
On Tue, Jun 2, 2015 at 1:35 PM, Till Rohrmann wrote:
> +1 for printOnTaskManager(prefix)
>
> On Tue, Jun 2, 2015 at 12:08 PM, Fabian Hueske wrote:
>
> > +1 for writeToWorkerStdOut(prefix)
> > On Jun 2, 2015 11:42, "Aljoscha Krettek" wrote:
> >
> > > +1 for pri
+1 for printOnTaskManager(prefix)
On Tue, Jun 2, 2015 at 12:08 PM, Fabian Hueske wrote:
> +1 for writeToWorkerStdOut(prefix)
> On Jun 2, 2015 11:42, "Aljoscha Krettek" wrote:
>
> > +1 for printOnTaskManager(prefix)
> >
> > On Tue, Jun 2, 2015 at 11:35 AM, Robert Metzger
> > wrote:
> > > I woul
Hi Florian,
I just wrote a patch for this problem. I wait until all tests pass and then
I’ll merge the fix. Thus, it will be included in the current master in the
late afternoon.
If you don’t want to wait that long, then you can also solve the issue with
.fromBreeze[org.apache.flink.ml.math.Vecto
Hi Flink Community,
we are implementing a KMeans algorithm in the ML part of Flink, but after
recent updates we ran into an issues with the vector to breeze converter.
We are getting the following compile error:
Error:(200, 75) ambiguous implicit values:
both value denseVectorConverter in obje
+1 for writeToWorkerStdOut(prefix)
On Jun 2, 2015 11:42, "Aljoscha Krettek" wrote:
> +1 for printOnTaskManager(prefix)
>
> On Tue, Jun 2, 2015 at 11:35 AM, Robert Metzger
> wrote:
> > I would like to reach consensus on this before the 0.9 release.
> >
> > So far we have the following ideas:
> >
+1 for printOnTaskManager(prefix)
On Tue, Jun 2, 2015 at 11:35 AM, Robert Metzger wrote:
> I would like to reach consensus on this before the 0.9 release.
>
> So far we have the following ideas:
>
> writeToWorkerStdOut(prefix)
> printOnTaskManager(prefix) (+1)
> logOnTaskManager(prefix)
>
> I'm a
I would like to reach consensus on this before the 0.9 release.
So far we have the following ideas:
writeToWorkerStdOut(prefix)
printOnTaskManager(prefix) (+1)
logOnTaskManager(prefix)
I'm against logOnTM because we are not logging the output, we are writing
or printing it.
*I would vote for d
Great to hear. This should no longer be a pain point once we support proper
cross validation.
On Tue, Jun 2, 2015 at 11:11 AM, Felix Neutatz
wrote:
> Yes, grid search solved the problem :)
>
> 2015-06-02 11:07 GMT+02:00 Till Rohrmann :
>
> > The SGD algorithm adapts the learning rate accordingly
Yes, grid search solved the problem :)
2015-06-02 11:07 GMT+02:00 Till Rohrmann :
> The SGD algorithm adapts the learning rate accordingly. However, this does
> not help if you choose the initial learning rate too large because then you
> calculate a weight vector in the first iterations from whi
@Robert
Thanks! I think the PR will be ready to merge soon :)
@Matthias
I fixed the finite-source issue on my branch, now every example and ITCase
runs and stopps without throwing an exception. Also, in case of finite
sources, the spout wrapper will not loop infinitely.
I will study your branch an
The SGD algorithm adapts the learning rate accordingly. However, this does
not help if you choose the initial learning rate too large because then you
calculate a weight vector in the first iterations from which it takes
really long to recover.
Cheer,
Till
On Mon, Jun 1, 2015 at 7:15 PM, Sachin G
Ufuk Celebi created FLINK-2132:
--
Summary: Java version parsing is not working for OpenJDK
Key: FLINK-2132
URL: https://issues.apache.org/jira/browse/FLINK-2132
Project: Flink
Issue Type: Bug
45 matches
Mail list logo
