Question: SourceFunction

Hi, I just encountered, that the (new streaming API) SourceFunction interface method cancel() states in the JavaDoc, that one might have a volatile field 'isRunning'. Why should this member be volatile? I do not see why different thread should access this field (I would claim, it will be

[jira] [Created] (FLINK-2233) InputFormat nextRecord Javadocs out of sync

Ufuk Celebi created FLINK-2233: -- Summary: InputFormat nextRecord Javadocs out of sync Key: FLINK-2233 URL: https://issues.apache.org/jira/browse/FLINK-2233 Project: Flink Issue Type: Bug

Re: Shading artifact name

We could replace the placeholder error with something more meaningful though... On Tue, Jun 16, 2015 at 1:27 PM, Márton Balassi balassi.mar...@gmail.com wrote: Ok, never mind. I got it: it is decided later dependent on the hadoop profile. Makes sense. On Tue, Jun 16, 2015 at 1:24 PM, Márton

[jira] [Created] (FLINK-2234) CoStreamWindow is Dependent on Input Ordering

Aljoscha Krettek created FLINK-2234: --- Summary: CoStreamWindow is Dependent on Input Ordering Key: FLINK-2234 URL: https://issues.apache.org/jira/browse/FLINK-2234 Project: Flink Issue

Re: Testing Apache Flink 0.9.0-rc2

I would like to include the fixes for the Table API: RowSerializer supports null values, Aggregations properly deal with null values. What do you think? On Mon, 15 Jun 2015 at 17:31 Aljoscha Krettek aljos...@apache.org wrote: I created this to help with release testing:

Re: Testing Apache Flink 0.9.0-rc2

@Max: The PR is good to go on my side. Does the job, could be a bit nicer though. Added to the document. On Tue, Jun 16, 2015 at 10:54 PM, Aljoscha Krettek aljos...@apache.org wrote: I added the two relevant Table API commits to the release doc. On Tue, 16 Jun 2015 at 21:49 Maximilian Michels

Re: Testing Apache Flink 0.9.0-rc2

I merged the legal and scheduler PR. On Tue, Jun 16, 2015 at 11:17 PM Márton Balassi balassi.mar...@gmail.com wrote: @Max: The PR is good to go on my side. Does the job, could be a bit nicer though. Added to the document. On Tue, Jun 16, 2015 at 10:54 PM, Aljoscha Krettek aljos...@apache.org

Re: Testing Apache Flink 0.9.0-rc2

I added the two relevant Table API commits to the release doc. On Tue, 16 Jun 2015 at 21:49 Maximilian Michels m...@apache.org wrote: +1 for adding the Table API fixes. @Till: It seems like you fixed the bug in the scheduler. Is https://github.com/apache/flink/pull/843 fixing it? @Marton:

[VOTE] Release Apache Flink 0.9.0 (release-0.9.0-rc3)

Dear community! After careful testing and three release candidates, I take a chance and hereby start a vote. Please vote on releasing the following candidate as Apache Flink version 0.9.0. - The commit to be voted on:

[jira] [Created] (FLINK-2231) Create a Serializer for Scala Enumerations

Stephan Ewen created FLINK-2231: --- Summary: Create a Serializer for Scala Enumerations Key: FLINK-2231 URL: https://issues.apache.org/jira/browse/FLINK-2231 Project: Flink Issue Type:

[jira] [Created] (FLINK-2230) Add Support for Null-Values in TupleSerializer

Shiti Saxena created FLINK-2230: --- Summary: Add Support for Null-Values in TupleSerializer Key: FLINK-2230 URL: https://issues.apache.org/jira/browse/FLINK-2230 Project: Flink Issue Type: Bug

Re: ClassNotFound exception on a reduce group (Cluster mode)

Hi, Thank you for your help. It was actually an issue between the Flink version I used to build my quickstart jar and the Flink version on the cluster. Cheers, Thomas Le 16/06/15 10:10, Till Rohrmann a écrit : I also tested your code with an additional sink and it is executed on a cluster

Re: Shading artifact name

Ok, never mind. I got it: it is decided later dependent on the hadoop profile. Makes sense. On Tue, Jun 16, 2015 at 1:24 PM, Márton Balassi balassi.mar...@gmail.com wrote: Hey, While reading through the flink-parent pom I have stumbled upon this. [1]