Re: [DISCUSS] Release Flink 1.5.3
Generally, for bugfix releases I only regard issues as blockers that were introduced in that version, which doesn't appear to be the case here. There's nothing speaking against another bugfix release soon after 1.5.3 once the issue is resolved. On 16.08.2018 18:46, Elias Levy wrote: If I may, think FLINK-10011 should be considered a blocker and included in 1.5.3 and a 1.4.3 release. On Thu, Aug 16, 2018 at 3:09 AM Dominik Wosiński wrote: +1, I agree that frequent releases are good for users. Dominik Wysłane z aplikacji Poczta dla Windows 10 Od: Stefan Richter Wysłano: czwartek, 16 sierpnia 2018 10:57 Do: dev Temat: Re: [DISCUSS] Release Flink 1.5.3 +1 sounds good. Stefan Am 16.08.2018 um 09:55 schrieb Piotr Nowojski : +1 from me Piotrek On 16 Aug 2018, at 09:40, Timo Walther wrote: Thank you for starting this discussion. +1 for this Regards, Timo Am 16.08.18 um 09:27 schrieb vino yang: Agree! This sounds very good. Till Rohrmann 于2018年8月16日周四 下午3:14写道: +1 for starting the release process 1.5.3 immediately. We can always create another bug fix release afterwards. I think the more often we release the better it is for our users, because they receive incremental improvements faster. Cheers, Till On Thu, Aug 16, 2018 at 8:52 AM Chesnay Schepler wrote: I would actually start with the release process today unless anyone objects. On 16.08.2018 08:46, vino yang wrote: Hi Chesnay, +1 I want to know when you plan to cut the branch. Thanks, vino. Chesnay Schepler 于2018年8月16日周四 下午2:29写道: Hello everyone, it has been a little over 2 weeks since 1.5.2 was released, and since then a number of fixes were added to the release-1.5 that I think we should push to users. Notable fixes, (among others) are FLINK-9969 which fixes a memory leak when using batch, FLINK-10066 that reduces the memory print on the JM for long-running jobs or FLINK-10070 that reverses a regression introduced in 1.5.2 due to which Flink could not be compiled with certain maven versions. I would of course volunteer as Release Manager.
Re: [DISCUSS] Release Flink 1.5.3
If I may, think FLINK-10011 should be considered a blocker and included in 1.5.3 and a 1.4.3 release. On Thu, Aug 16, 2018 at 3:09 AM Dominik Wosiński wrote: > +1, I agree that frequent releases are good for users. > Dominik > > > Wysłane z aplikacji Poczta dla Windows 10 > > Od: Stefan Richter > Wysłano: czwartek, 16 sierpnia 2018 10:57 > Do: dev > Temat: Re: [DISCUSS] Release Flink 1.5.3 > > +1 sounds good. > > Stefan > > > Am 16.08.2018 um 09:55 schrieb Piotr Nowojski : > > > > +1 from me > > > > Piotrek > > > >> On 16 Aug 2018, at 09:40, Timo Walther wrote: > >> > >> Thank you for starting this discussion. > >> > >> +1 for this > >> > >> Regards, > >> Timo > >> > >> Am 16.08.18 um 09:27 schrieb vino yang: > >>> Agree! This sounds very good. > >>> > >>> Till Rohrmann 于2018年8月16日周四 下午3:14写道: > >>> > +1 for starting the release process 1.5.3 immediately. We can always > create another bug fix release afterwards. I think the more often we > release the better it is for our users, because they receive > incremental > improvements faster. > > Cheers, > Till > > On Thu, Aug 16, 2018 at 8:52 AM Chesnay Schepler > wrote: > > > I would actually start with the release process today unless anyone > > objects. > > > > On 16.08.2018 08:46, vino yang wrote: > >> Hi Chesnay, > >> > >> +1 > >> I want to know when you plan to cut the branch. > >> > >> Thanks, vino. > >> > >> Chesnay Schepler 于2018年8月16日周四 下午2:29写道: > >> > >>> Hello everyone, > >>> > >>> it has been a little over 2 weeks since 1.5.2 was released, and > since > >>> then a number of fixes were added to the release-1.5 that I think > we > >>> should push to users. > >>> > >>> Notable fixes, (among others) are FLINK-9969 which fixes a memory > leak > >>> when using batch, FLINK-10066 that reduces the memory print on the > JM > >>> for long-running jobs or FLINK-10070 that reverses a regression > >>> introduced in 1.5.2 due to which Flink could not be compiled with > >>> certain maven versions. > >>> > >>> I would of course volunteer as Release Manager. > >>> > >>> > > > >> > > > > >
[jira] [Created] (FLINK-10164) Add support for resuming from savepoints to StandaloneJobClusterEntrypoint
Till Rohrmann created FLINK-10164: - Summary: Add support for resuming from savepoints to StandaloneJobClusterEntrypoint Key: FLINK-10164 URL: https://issues.apache.org/jira/browse/FLINK-10164 Project: Flink Issue Type: Improvement Components: Distributed Coordination Affects Versions: 1.6.0, 1.7.0 Reporter: Till Rohrmann Assignee: Till Rohrmann Fix For: 1.6.1, 1.7.0 The {{StandaloneJobClusterEntrypoint}} should support to resume from a savepoint/checkpoint. I suggest to introduce an optional command line parameter for specifying the savepoint/checkpoint path. -- This message was sent by Atlassian JIRA (v7.6.3#76005)
Re: How to handle tests that need docker?
I agree with Dominik. We could add this test to `flink-end-to-end-test/run-pre-commit-tests.sh` or `flink-end-to-end-test/run-nightly-tests.sh` like we did for the `flink-end-to-end-test/test-scripts/test_docker_embedded_job.sh` and `flinke-end-to-end-test/test-scripts/test_yarn_kerberos_docker.sh`. Cheers, Till On Thu, Aug 16, 2018 at 2:10 PM Dominik Wosiński wrote: > Imho, > Such tests should be off by default as they impose some requirements on > the people trying to build Flink from source. And one would have to install > docker only for the tests to pass. But of course they should be enabled for > CI builds. > > Best Regards, > Dominik. > > Wysłane z aplikacji Poczta dla Windows 10 > > Od: Niels Basjes > Wysłano: czwartek, 16 sierpnia 2018 14:05 > Do: dev@flink.apache.org > Temat: How to handle tests that need docker? > > Hi, > > I'm working together with Richard Deurwaarder on the PubSub Source and Sink > https://issues.apache.org/jira/browse/FLINK-9311 > > We currently have most of the code done and we also have several tests that > verify the behavior by actually reading and writing messages to and from > PubSub. > These tests use a docker image provided by Google that includes the pubsub > emulator. > > Starting and stopping this thing is really fast. > Starting up, creating topic, running a few tests and cleaning up takes > about 5 seconds. > > Although the complete set of tests run on my machine in about 20-25 seconds > it does mean that anyone doing that MUST have docker installed locally. > > So my question to the committers: Do we simply leave these tests "always > on" or do you want them to be "off by default" or something else. > I case you do not want the "always on" then please indicate how we should > implement the "on"/"off" switch for these tests. > > Thanks. > > -- > Best regards / Met vriendelijke groeten, > > Niels Basjes > >
[jira] [Created] (FLINK-10163) Support CREATE VIEW in SQL Client
Timo Walther created FLINK-10163: Summary: Support CREATE VIEW in SQL Client Key: FLINK-10163 URL: https://issues.apache.org/jira/browse/FLINK-10163 Project: Flink Issue Type: New Feature Components: Table API SQL Reporter: Timo Walther Assignee: Timo Walther The possibility to define a name for a subquery would improve the usability of the SQL Client. The SQL standard defines \{{CREATE VIEW}} for defining a virtual table. Example: {code} CREATE VIEW view_name AS SELECT {code} -- This message was sent by Atlassian JIRA (v7.6.3#76005)
[jira] [Created] (FLINK-10162) Add host in checkpoint detail
xymaqingxiang created FLINK-10162: - Summary: Add host in checkpoint detail Key: FLINK-10162 URL: https://issues.apache.org/jira/browse/FLINK-10162 Project: Flink Issue Type: Improvement Reporter: xymaqingxiang Assignee: xymaqingxiang Add the host information to the checkpoint so that if the checkpoint fails, you can specifically login to the appropriate machine to see the cause of the failure. -- This message was sent by Atlassian JIRA (v7.6.3#76005)
[jira] [Created] (FLINK-10161) Add host in checkpoint detail
xymaqingxiang created FLINK-10161: - Summary: Add host in checkpoint detail Key: FLINK-10161 URL: https://issues.apache.org/jira/browse/FLINK-10161 Project: Flink Issue Type: New Feature Reporter: xymaqingxiang Add the host information to the checkpoint so that if the checkpoint fails, you can specifically login to the appropriate machine to see the cause of the failure. -- This message was sent by Atlassian JIRA (v7.6.3#76005)
[jira] [Created] (FLINK-10160) Index out of bound when serializing operator state
Zdenek created FLINK-10160: -- Summary: Index out of bound when serializing operator state Key: FLINK-10160 URL: https://issues.apache.org/jira/browse/FLINK-10160 Project: Flink Issue Type: Bug Components: Core Affects Versions: 1.4.2 Environment: Cloudera Hadoop 2.6.5 JVM: Java HotSpot(TM) 64-Bit Server VM - Oracle Corporation - 1.8/25.121-b13 Scala 2.11 Reporter: Zdenek Attachments: log.txt I am getting unexpected randomly happend error when checkpoint state is serialized to state backend (In-Memory). Source code usage: {code:java} env.setStreamTimeCharacteristic(TimeCharacteristic.EventTime) env.registerType(classOf[ImpressionMetric]) env.registerType(classOf[RecommendationMetric]) env.registerType(classOf[AdRequestMetric]) env.registerType(classOf[AdRequestStatusMetric]) val impressionsStream = env .addSource(impsSource) .map(imp => ImpressionMetric(imp)) .map(imp => imp.asInstanceOf[DomainMetric]) .name("Impressions") val recommendationsStream = env .addSource(recommendationsSource) .map(rcmd => RecommendationMetric(rcmd)) .map(rcmd => rcmd.asInstanceOf[DomainMetric]) .name("Recommendations") val adRequestsStream = env .addSource(adRequestsSource) .flatMap(new MapToAdRequestDomainMetrics()) .name("Ad requests") {code} Log with error, longer version in [^log.txt] {code:java} 2018-08-16 12:49:37,663 INFO org.apache.flink.runtime.checkpoint.CheckpointCoordinator - Triggering checkpoint 1 @ 1534416577551 2018-08-16 12:49:38,100 INFO org.apache.flink.runtime.checkpoint.CheckpointCoordinator - Completed checkpoint 1 (1404248 bytes in 545 ms). 2018-08-16 12:57:54,217 INFO org.apache.flink.runtime.checkpoint.CheckpointCoordinator - Triggering checkpoint 2 @ 1534417074216 2018-08-16 12:57:54,368 INFO org.apache.flink.runtime.checkpoint.CheckpointCoordinator - Completed checkpoint 2 (2199656 bytes in 152 ms). 2018-08-16 13:07:54,217 INFO org.apache.flink.runtime.checkpoint.CheckpointCoordinator - Triggering checkpoint 3 @ 1534417674216 2018-08-16 13:17:54,218 INFO org.apache.flink.runtime.checkpoint.CheckpointCoordinator - Checkpoint 3 expired before completing. 2018-08-16 13:17:54,220 INFO org.apache.flink.runtime.checkpoint.CheckpointCoordinator - Triggering checkpoint 4 @ 1534418274220 2018-08-16 13:27:54,220 INFO org.apache.flink.runtime.checkpoint.CheckpointCoordinator - Checkpoint 4 expired before completing. 2018-08-16 13:27:54,222 INFO org.apache.flink.runtime.checkpoint.CheckpointCoordinator - Triggering checkpoint 5 @ 1534418874221 2018-08-16 13:29:36,640 WARN org.apache.flink.runtime.checkpoint.CheckpointCoordinator - Received late message for now expired checkpoint attempt 3 from e4ce857f2ef5dd4cc75a48b6fdb7b694 of job a98b29dcef6ad8eac28f8290034a0590. 2018-08-16 13:29:36,642 WARN org.apache.flink.runtime.checkpoint.CheckpointCoordinator - Received late message for now expired checkpoint attempt 3 from 1db1e9c995bae21d6c81e828e11f50a6 of job a98b29dcef6ad8eac28f8290034a0590. 2018-08-16 13:29:36,661 WARN org.apache.flink.runtime.checkpoint.CheckpointCoordinator - Received late message for now expired checkpoint attempt 4 from 0f62f47cdd04acf58cf8aa6c11288950 of job a98b29dcef6ad8eac28f8290034a0590. 2018-08-16 13:29:36,662 WARN org.apache.flink.runtime.checkpoint.CheckpointCoordinator - Received late message for now expired checkpoint attempt 4 from e4ce857f2ef5dd4cc75a48b6fdb7b694 of job a98b29dcef6ad8eac28f8290034a0590. 2018-08-16 13:29:36,663 WARN org.apache.flink.runtime.checkpoint.CheckpointCoordinator - Received late message for now expired checkpoint attempt 3 from 0f62f47cdd04acf58cf8aa6c11288950 of job a98b29dcef6ad8eac28f8290034a0590. 2018-08-16 13:29:36,967 INFO org.apache.flink.runtime.executiongraph.ExecutionGraph - Source: Custom Source -> Ad requests (1/2) (1db1e9c995bae21d6c81e828e11f50a6) switched from RUNNING to FAILED. java.lang.Exception: Error while triggering checkpoint 4 for Source: Custom Source -> Ad requests (1/2) at org.apache.flink.runtime.taskmanager.Task$2.run(Task.java:1210) at java.util.concurrent.Executors$RunnableAdapter.call(Executors.java:511) at java.util.concurrent.FutureTask.run(FutureTask.java:266) at java.util.concurrent.ThreadPoolExecutor.runWorker(ThreadPoolExecutor.java:1142) at java.util.concurrent.ThreadPoolExecutor$Worker.run(ThreadPoolExecutor.java:617) at java.lang.Thread.run(Thread.java:745) Caused by: java.lang.Exception: Could not perform checkpoint 4 for operator Source: Custom Source -> Ad requests (1/2). at org.apache.flink.streaming.runtime.tasks.StreamTask.triggerCheckpoint(StreamTask.java:544) at org.apache.flink.streaming.runtime.tasks.SourceStreamTask.triggerCheckpoint(SourceStreamTask.java:111) at org.apache.flink.runtime.taskmanager.Task$2.run(Task.java:1199) ... 5 more Caused by: java.lang.Exception:
ODP: How to handle tests that need docker?
Imho, Such tests should be off by default as they impose some requirements on the people trying to build Flink from source. And one would have to install docker only for the tests to pass. But of course they should be enabled for CI builds. Best Regards, Dominik. Wysłane z aplikacji Poczta dla Windows 10 Od: Niels Basjes Wysłano: czwartek, 16 sierpnia 2018 14:05 Do: dev@flink.apache.org Temat: How to handle tests that need docker? Hi, I'm working together with Richard Deurwaarder on the PubSub Source and Sink https://issues.apache.org/jira/browse/FLINK-9311 We currently have most of the code done and we also have several tests that verify the behavior by actually reading and writing messages to and from PubSub. These tests use a docker image provided by Google that includes the pubsub emulator. Starting and stopping this thing is really fast. Starting up, creating topic, running a few tests and cleaning up takes about 5 seconds. Although the complete set of tests run on my machine in about 20-25 seconds it does mean that anyone doing that MUST have docker installed locally. So my question to the committers: Do we simply leave these tests "always on" or do you want them to be "off by default" or something else. I case you do not want the "always on" then please indicate how we should implement the "on"/"off" switch for these tests. Thanks. -- Best regards / Met vriendelijke groeten, Niels Basjes
How to handle tests that need docker?
Hi, I'm working together with Richard Deurwaarder on the PubSub Source and Sink https://issues.apache.org/jira/browse/FLINK-9311 We currently have most of the code done and we also have several tests that verify the behavior by actually reading and writing messages to and from PubSub. These tests use a docker image provided by Google that includes the pubsub emulator. Starting and stopping this thing is really fast. Starting up, creating topic, running a few tests and cleaning up takes about 5 seconds. Although the complete set of tests run on my machine in about 20-25 seconds it does mean that anyone doing that MUST have docker installed locally. So my question to the committers: Do we simply leave these tests "always on" or do you want them to be "off by default" or something else. I case you do not want the "always on" then please indicate how we should implement the "on"/"off" switch for these tests. Thanks. -- Best regards / Met vriendelijke groeten, Niels Basjes
[jira] [Created] (FLINK-10159) TestHarness#initializeState(xyz) calls after TestHarness#open() are being silently ignored
Piotr Nowojski created FLINK-10159: -- Summary: TestHarness#initializeState(xyz) calls after TestHarness#open() are being silently ignored Key: FLINK-10159 URL: https://issues.apache.org/jira/browse/FLINK-10159 Project: Flink Issue Type: Bug Components: Tests Affects Versions: 1.6.0 Reporter: Piotr Nowojski This is an old issue. Incorrect order of initializeState and open result to initializeState being ignored. For example in this code: {code:java} testHarness = createTestHarness(topic); testHarness.setup(); testHarness.open(); testHarness.initializeState(snapshot1); {code} Which is miss-leading both for Flink developers and for users (since we recommend using test harness for unit tests). -- This message was sent by Atlassian JIRA (v7.6.3#76005)
[VOTE] Release 1.5.3, release candidate #1
Hi everyone, Please review and vote on the release candidate #1 for the version 1.5.3, as follows: [ ] +1, Approve the release [ ] -1, Do not approve the release (please provide specific comments) The complete staging area is available for your review, which includes: * JIRA release notes [1], * the official Apache source release and binary convenience releases to be deployed to dist.apache.org [2], which are signed with the key with fingerprint 11D464BA [3], * all artifacts to be deployed to the Maven Central Repository [4], * source code tag "release-1.5.3-rc1" [5], * website pull request listing the new release and adding announcement blog post [6]. The vote will be open for at least 72 hours. It is adopted by majority approval, with at least 3 PMC affirmative votes. Thanks, Chesnay [1] https://issues.apache.org/jira/secure/ReleaseNote.jspa?projectId=12315522=12343777 [2] https://dist.apache.org/repos/dist/dev/flink/flink-1.5.3/ [3] https://dist.apache.org/repos/dist/release/flink/KEYS [4] https://repository.apache.org/content/repositories/orgapacheflink-1179 [5] https://gitbox.apache.org/repos/asf?p=flink.git;a=tag;h=refs/tags/release-1.5.3-rc1 [6] https://github.com/apache/flink-web/pull/119
ODP: [DISCUSS] Release Flink 1.5.3
+1, I agree that frequent releases are good for users. Dominik Wysłane z aplikacji Poczta dla Windows 10 Od: Stefan Richter Wysłano: czwartek, 16 sierpnia 2018 10:57 Do: dev Temat: Re: [DISCUSS] Release Flink 1.5.3 +1 sounds good. Stefan > Am 16.08.2018 um 09:55 schrieb Piotr Nowojski : > > +1 from me > > Piotrek > >> On 16 Aug 2018, at 09:40, Timo Walther wrote: >> >> Thank you for starting this discussion. >> >> +1 for this >> >> Regards, >> Timo >> >> Am 16.08.18 um 09:27 schrieb vino yang: >>> Agree! This sounds very good. >>> >>> Till Rohrmann 于2018年8月16日周四 下午3:14写道: >>> +1 for starting the release process 1.5.3 immediately. We can always create another bug fix release afterwards. I think the more often we release the better it is for our users, because they receive incremental improvements faster. Cheers, Till On Thu, Aug 16, 2018 at 8:52 AM Chesnay Schepler wrote: > I would actually start with the release process today unless anyone > objects. > > On 16.08.2018 08:46, vino yang wrote: >> Hi Chesnay, >> >> +1 >> I want to know when you plan to cut the branch. >> >> Thanks, vino. >> >> Chesnay Schepler 于2018年8月16日周四 下午2:29写道: >> >>> Hello everyone, >>> >>> it has been a little over 2 weeks since 1.5.2 was released, and since >>> then a number of fixes were added to the release-1.5 that I think we >>> should push to users. >>> >>> Notable fixes, (among others) are FLINK-9969 which fixes a memory leak >>> when using batch, FLINK-10066 that reduces the memory print on the JM >>> for long-running jobs or FLINK-10070 that reverses a regression >>> introduced in 1.5.2 due to which Flink could not be compiled with >>> certain maven versions. >>> >>> I would of course volunteer as Release Manager. >>> >>> > >> >
Re: [DISCUSS] Release Flink 1.5.3
+1 sounds good. Stefan > Am 16.08.2018 um 09:55 schrieb Piotr Nowojski : > > +1 from me > > Piotrek > >> On 16 Aug 2018, at 09:40, Timo Walther wrote: >> >> Thank you for starting this discussion. >> >> +1 for this >> >> Regards, >> Timo >> >> Am 16.08.18 um 09:27 schrieb vino yang: >>> Agree! This sounds very good. >>> >>> Till Rohrmann 于2018年8月16日周四 下午3:14写道: >>> +1 for starting the release process 1.5.3 immediately. We can always create another bug fix release afterwards. I think the more often we release the better it is for our users, because they receive incremental improvements faster. Cheers, Till On Thu, Aug 16, 2018 at 8:52 AM Chesnay Schepler wrote: > I would actually start with the release process today unless anyone > objects. > > On 16.08.2018 08:46, vino yang wrote: >> Hi Chesnay, >> >> +1 >> I want to know when you plan to cut the branch. >> >> Thanks, vino. >> >> Chesnay Schepler 于2018年8月16日周四 下午2:29写道: >> >>> Hello everyone, >>> >>> it has been a little over 2 weeks since 1.5.2 was released, and since >>> then a number of fixes were added to the release-1.5 that I think we >>> should push to users. >>> >>> Notable fixes, (among others) are FLINK-9969 which fixes a memory leak >>> when using batch, FLINK-10066 that reduces the memory print on the JM >>> for long-running jobs or FLINK-10070 that reverses a regression >>> introduced in 1.5.2 due to which Flink could not be compiled with >>> certain maven versions. >>> >>> I would of course volunteer as Release Manager. >>> >>> > >> >
[jira] [Created] (FLINK-10158) The DataOutputSerializer may consume excessive memory
aitozi created FLINK-10158: -- Summary: The DataOutputSerializer may consume excessive memory Key: FLINK-10158 URL: https://issues.apache.org/jira/browse/FLINK-10158 Project: Flink Issue Type: Improvement Components: Core Affects Versions: 1.6.0, 1.5.2, 1.4.2 Reporter: aitozi Assignee: aitozi I found that the dataOutputSerializer clear the intermediate buffer when the buffer exceed the 5M (as a fixed configuration), But when we encountered the rebalance or keyByPartition and the downstream has a large parallel it will also consume a lot memory, we can do two things : 1. make this config configurable 2. like the https://issues.apache.org/jira/projects/FLINK/issues/FLINK-1326 mentioned, we can make the serializer one for the output What's your idea ? [~StephanEwen] -- This message was sent by Atlassian JIRA (v7.6.3#76005)
[jira] [Created] (FLINK-10157) The TTL state not allow `null` state value
chengjie.wu created FLINK-10157: --- Summary: The TTL state not allow `null` state value Key: FLINK-10157 URL: https://issues.apache.org/jira/browse/FLINK-10157 Project: Flink Issue Type: Bug Components: State Backends, Checkpointing Affects Versions: 1.6.0 Environment: Flink:1.6.0 Scala:2.11 JDK:1.8 Reporter: chengjie.wu Attachments: StateWithOutTtlTest.scala, StateWithTtlTest.scala In the previous version or when StateTtl is not enabled,MapState allows `null` value,that means after {code:java} mapState.put("key", null){code} , then {code:java} mapState.contains("key"){code} will return {color:#FF}*true*{color}, but when StateTtl is enabled, {code:java} mapState.contains("key"){code} will return {color:#FF}*false*{color}(*the key has not expired*). So I think the field `userValue` in `org.apache.flink.runtime.state.ttl.TtlValue` should allow `null` value. User state is null may not means the TtlValue should be null. {code:java} /** * This class wraps user value of state with TTL. * * @param Type of the user value of state with TTL */ class TtlValue implements Serializable { private final T userValue; private final long lastAccessTimestamp; TtlValue(T userValue, long lastAccessTimestamp) { Preconditions.checkNotNull(userValue); this.userValue = userValue; this.lastAccessTimestamp = lastAccessTimestamp; } T getUserValue() { return userValue; } long getLastAccessTimestamp() { return lastAccessTimestamp; } } {code} Am I understanding right? This is my test class. [^StateWithTtlTest.scala] [^StateWithOutTtlTest.scala] ^Thanks!:)^ -- This message was sent by Atlassian JIRA (v7.6.3#76005)
Re: [DISCUSS] Release Flink 1.5.3
+1 from me Piotrek > On 16 Aug 2018, at 09:40, Timo Walther wrote: > > Thank you for starting this discussion. > > +1 for this > > Regards, > Timo > > Am 16.08.18 um 09:27 schrieb vino yang: >> Agree! This sounds very good. >> >> Till Rohrmann 于2018年8月16日周四 下午3:14写道: >> >>> +1 for starting the release process 1.5.3 immediately. We can always >>> create another bug fix release afterwards. I think the more often we >>> release the better it is for our users, because they receive incremental >>> improvements faster. >>> >>> Cheers, >>> Till >>> >>> On Thu, Aug 16, 2018 at 8:52 AM Chesnay Schepler >>> wrote: >>> I would actually start with the release process today unless anyone objects. On 16.08.2018 08:46, vino yang wrote: > Hi Chesnay, > > +1 > I want to know when you plan to cut the branch. > > Thanks, vino. > > Chesnay Schepler 于2018年8月16日周四 下午2:29写道: > >> Hello everyone, >> >> it has been a little over 2 weeks since 1.5.2 was released, and since >> then a number of fixes were added to the release-1.5 that I think we >> should push to users. >> >> Notable fixes, (among others) are FLINK-9969 which fixes a memory leak >> when using batch, FLINK-10066 that reduces the memory print on the JM >> for long-running jobs or FLINK-10070 that reverses a regression >> introduced in 1.5.2 due to which Flink could not be compiled with >> certain maven versions. >> >> I would of course volunteer as Release Manager. >> >> >
Re: [DISCUSS] Release Flink 1.5.3
Thank you for starting this discussion. +1 for this Regards, Timo Am 16.08.18 um 09:27 schrieb vino yang: Agree! This sounds very good. Till Rohrmann 于2018年8月16日周四 下午3:14写道: +1 for starting the release process 1.5.3 immediately. We can always create another bug fix release afterwards. I think the more often we release the better it is for our users, because they receive incremental improvements faster. Cheers, Till On Thu, Aug 16, 2018 at 8:52 AM Chesnay Schepler wrote: I would actually start with the release process today unless anyone objects. On 16.08.2018 08:46, vino yang wrote: Hi Chesnay, +1 I want to know when you plan to cut the branch. Thanks, vino. Chesnay Schepler 于2018年8月16日周四 下午2:29写道: Hello everyone, it has been a little over 2 weeks since 1.5.2 was released, and since then a number of fixes were added to the release-1.5 that I think we should push to users. Notable fixes, (among others) are FLINK-9969 which fixes a memory leak when using batch, FLINK-10066 that reduces the memory print on the JM for long-running jobs or FLINK-10070 that reverses a regression introduced in 1.5.2 due to which Flink could not be compiled with certain maven versions. I would of course volunteer as Release Manager.
Re: [DISCUSS] Release Flink 1.5.3
Agree! This sounds very good. Till Rohrmann 于2018年8月16日周四 下午3:14写道: > +1 for starting the release process 1.5.3 immediately. We can always > create another bug fix release afterwards. I think the more often we > release the better it is for our users, because they receive incremental > improvements faster. > > Cheers, > Till > > On Thu, Aug 16, 2018 at 8:52 AM Chesnay Schepler > wrote: > >> I would actually start with the release process today unless anyone >> objects. >> >> On 16.08.2018 08:46, vino yang wrote: >> > Hi Chesnay, >> > >> > +1 >> > I want to know when you plan to cut the branch. >> > >> > Thanks, vino. >> > >> > Chesnay Schepler 于2018年8月16日周四 下午2:29写道: >> > >> >> Hello everyone, >> >> >> >> it has been a little over 2 weeks since 1.5.2 was released, and since >> >> then a number of fixes were added to the release-1.5 that I think we >> >> should push to users. >> >> >> >> Notable fixes, (among others) are FLINK-9969 which fixes a memory leak >> >> when using batch, FLINK-10066 that reduces the memory print on the JM >> >> for long-running jobs or FLINK-10070 that reverses a regression >> >> introduced in 1.5.2 due to which Flink could not be compiled with >> >> certain maven versions. >> >> >> >> I would of course volunteer as Release Manager. >> >> >> >> >> >>
Re: [DISCUSS] Release Flink 1.5.3
+1 for starting the release process 1.5.3 immediately. We can always create another bug fix release afterwards. I think the more often we release the better it is for our users, because they receive incremental improvements faster. Cheers, Till On Thu, Aug 16, 2018 at 8:52 AM Chesnay Schepler wrote: > I would actually start with the release process today unless anyone > objects. > > On 16.08.2018 08:46, vino yang wrote: > > Hi Chesnay, > > > > +1 > > I want to know when you plan to cut the branch. > > > > Thanks, vino. > > > > Chesnay Schepler 于2018年8月16日周四 下午2:29写道: > > > >> Hello everyone, > >> > >> it has been a little over 2 weeks since 1.5.2 was released, and since > >> then a number of fixes were added to the release-1.5 that I think we > >> should push to users. > >> > >> Notable fixes, (among others) are FLINK-9969 which fixes a memory leak > >> when using batch, FLINK-10066 that reduces the memory print on the JM > >> for long-running jobs or FLINK-10070 that reverses a regression > >> introduced in 1.5.2 due to which Flink could not be compiled with > >> certain maven versions. > >> > >> I would of course volunteer as Release Manager. > >> > >> > >
Re: [DISCUSS] Release Flink 1.5.3
I would actually start with the release process today unless anyone objects. On 16.08.2018 08:46, vino yang wrote: Hi Chesnay, +1 I want to know when you plan to cut the branch. Thanks, vino. Chesnay Schepler 于2018年8月16日周四 下午2:29写道: Hello everyone, it has been a little over 2 weeks since 1.5.2 was released, and since then a number of fixes were added to the release-1.5 that I think we should push to users. Notable fixes, (among others) are FLINK-9969 which fixes a memory leak when using batch, FLINK-10066 that reduces the memory print on the JM for long-running jobs or FLINK-10070 that reverses a regression introduced in 1.5.2 due to which Flink could not be compiled with certain maven versions. I would of course volunteer as Release Manager.
Re: [DISCUSS] Release Flink 1.5.3
Hi Chesnay, +1 I want to know when you plan to cut the branch. Thanks, vino. Chesnay Schepler 于2018年8月16日周四 下午2:29写道: > Hello everyone, > > it has been a little over 2 weeks since 1.5.2 was released, and since > then a number of fixes were added to the release-1.5 that I think we > should push to users. > > Notable fixes, (among others) are FLINK-9969 which fixes a memory leak > when using batch, FLINK-10066 that reduces the memory print on the JM > for long-running jobs or FLINK-10070 that reverses a regression > introduced in 1.5.2 due to which Flink could not be compiled with > certain maven versions. > > I would of course volunteer as Release Manager. > >
[DISCUSS] Release Flink 1.5.3
Hello everyone, it has been a little over 2 weeks since 1.5.2 was released, and since then a number of fixes were added to the release-1.5 that I think we should push to users. Notable fixes, (among others) are FLINK-9969 which fixes a memory leak when using batch, FLINK-10066 that reduces the memory print on the JM for long-running jobs or FLINK-10070 that reverses a regression introduced in 1.5.2 due to which Flink could not be compiled with certain maven versions. I would of course volunteer as Release Manager.