答复：UpdateTaskExecutionState during JobManager failover

Hi Stephan,  Thank you for detail explaination.  As you said, my opition is to keep task still running druing jobmanager failover, even though sending update status failed. For the first reason you mentioned, if i understand correctly, the key issue is status out of sync between taskmanager and

Re: FileNotFoundException thrown by BlobCache when running "mvn test" against flink-runtime 0.10 for Scala 2.11

My bad. Gist with exception is here: https://gist.github.com/revprez/5a730a45674f0fc6e52b Prez Cannady p: 617 500 3378 e: revp...@opencorrelate.org GH: https://github.com/opencorrelate

FileNotFoundException thrown by BlobCache when running "mvn test" against flink-runtime 0.10 for Scala 2.11

Steps Forked, cut bug fix branch off of release–0.10. One line change to fix class loader issue. Ran mvn clean test in flink-runtime. All 894 tests pass. Ran ./tools/change-scala-version 2.11 in the root. Ran mvn clean test -Dscala.version=2.11.7 in flink-runtime. Expected result All 894 tests pa

flink-shaded-hadoop modules disconnected from parent pom

Hi folks, I noticed today that the parent pom for the flink-shaded-hadoop pom (and thus also it's children) are not using ${ROOT}/pom.xml as their parent. However, ${ROOT}/pom.xml lists the hierarchy as a module. I'm curious to know why this is. It seems one artifact of this disconnect is that too

[jira] [Created] (FLINK-3237) Create AsynchronousKvStateSnapshot similar to AsynchronousStateHandle

Gyula Fora created FLINK-3237: - Summary: Create AsynchronousKvStateSnapshot similar to AsynchronousStateHandle Key: FLINK-3237 URL: https://issues.apache.org/jira/browse/FLINK-3237 Project: Flink

[jira] [Created] (FLINK-3236) Flink user code classloader should have Flink classloader as parent classloader

Stephan Ewen created FLINK-3236: --- Summary: Flink user code classloader should have Flink classloader as parent classloader Key: FLINK-3236 URL: https://issues.apache.org/jira/browse/FLINK-3236 Project:

Re: Dripping the Flink-on-Tez code for Flink 1.0

I’ve created an issue [1] and closed all other Tez issues. [1] https://issues.apache.org/jira/browse/FLINK-3235 > On 11 Jan 2016, at 02:52, Henry Saputra wrote: > > +1 > > I am always for simplifying our code base when possible. > > > > On Sunday, January 10, 2016, Stephan Ewen wrote: > >

[jira] [Created] (FLINK-3235) Drop Flink-on-Tez code

Ufuk Celebi created FLINK-3235: -- Summary: Drop Flink-on-Tez code Key: FLINK-3235 URL: https://issues.apache.org/jira/browse/FLINK-3235 Project: Flink Issue Type: Improvement Components

Re: [DISCUSS] Remove Combinable Annotation from DataSet API

+1 for approach #1 - Henry On Thu, Jan 14, 2016 at 5:35 AM, Chiwan Park wrote: > I’m also for approach #1. Now is nice time to apply some API-breaking > changes. > > > On Jan 14, 2016, at 1:19 AM, Aljoscha Krettek > wrote: > > > > I’m also for Approach #1. I like simplifying things. > >> On

Re: [DISCUSS] Remove Combinable Annotation from DataSet API

I’m also for approach #1. Now is nice time to apply some API-breaking changes. > On Jan 14, 2016, at 1:19 AM, Aljoscha Krettek wrote: > > I’m also for Approach #1. I like simplifying things. >> On 13 Jan 2016, at 14:25, Vasiliki Kalavri wrote: >> >> Hi, >> >> ​+1 for removing the Combinable a

[jira] [Created] (FLINK-3234) SortPartition does not support KeySelectorFunctions

Fabian Hueske created FLINK-3234: Summary: SortPartition does not support KeySelectorFunctions Key: FLINK-3234 URL: https://issues.apache.org/jira/browse/FLINK-3234 Project: Flink Issue Type:

[jira] [Created] (FLINK-3233) PartitionOperator does not support expression keys on atomic types

Fabian Hueske created FLINK-3233: Summary: PartitionOperator does not support expression keys on atomic types Key: FLINK-3233 URL: https://issues.apache.org/jira/browse/FLINK-3233 Project: Flink

Re: UpdateTaskExecutionState during JobManager failover

Hi! That is a super interesting idea. If I understand you correctly, you are suggesting to try and reconcile the TaskManagers and the JobManager before restarting the job. That would mean that in case of a master failure, the jobs may simply continue to run. That would be a nice enhancements, but

Re: [DISCUSS] Git force pushing and deletion of branchs

+1 for protecting all branches including master. > On Jan 14, 2016, at 1:20 AM, Aljoscha Krettek wrote: > > +1 on protecting the master >> On 13 Jan 2016, at 14:46, Márton Balassi wrote: >> >> +1 >> >> On Wed, Jan 13, 2016 at 12:37 PM, Matthias J. Sax wrote: >> >>> +1 >>> >>> On 01/13/2016

UpdateTaskExecutionState during JobManager failover

Hi,     As i know, when TaskManager send UpdateTaskExecutionState to JobManager, if the JobManager failover and the future response is fail, the task will be failed. Is it feasible to retry send UpdateTaskExecutionState again when future response fail until success. In JobManager HA mode, the  U