I don't know for sure. Stephan worked on this feature. I think the behavior / log messages are misleading and confusing right now.
On Fri, Jun 2, 2017 at 12:28 PM, Tzu-Li (Gordon) Tai <tzuli...@apache.org> wrote: > Should https://issues.apache.org/jira/browse/FLINK-6643 (Flink restarts > job in HA even if NoRestartStrategy is set) also be included? > > > On 2 June 2017 at 11:21:33 AM, Robert Metzger (rmetz...@apache.org) wrote: > > I agree with both! > > I've attached a label to all JIRAs for the 1.3.1 release. Once this list > has 0 open issues, I'll create 1.3.1: > https://issues.apache.org/jira/issues/?jql=labels%20%3D% > 20flink-rel-1.3.1-blockers > > On Fri, Jun 2, 2017 at 11:14 AM, Chesnay Schepler <ches...@apache.org> > wrote: > > > We should give a good error message when the state migration fails as > > described in FLINK-6742. > > > > > > On 02.06.2017 11:05, Robert Metzger wrote: > > > >> Hi devs, > >> > >> I would like to release Apache Flink 1.3.1 with the following fixes: > >> > >> - FLINK-6812 Elasticsearch 5 release artifacts not published to Maven > >> central > >> - FLINK-6783 Wrongly extracted TypeInformations for > >> WindowedStream::aggregate > >> - FLINK-6780 ExternalTableSource should add time attributes in the row > >> type > >> - FLINK-6775 StateDescriptor cannot be shared by multiple subtasks > >> - FLINK-6763 Inefficient PojoSerializerConfigSnapshot serialization > format > >> - FLINK-6764 Deduplicate stateless TypeSerializers when serializing > >> composite TypeSerializers > >> > >> Is there anything else that we need to wait for before we vote on the > >> first > >> RC? > >> > >> > >> Regards, > >> Robert > >> > >> > > >
