Rong Rong created FLINK-11271:
-
Summary: Improve Kerberos Credential Distribution
Key: FLINK-11271
URL: https://issues.apache.org/jira/browse/FLINK-11271
Project: Flink
Issue Type: Improvement
Chesnay Schepler created FLINK-11270:
Summary: Do not include hadoop in flink-dist by default
Key: FLINK-11270
URL: https://issues.apache.org/jira/browse/FLINK-11270
Project: Flink
Issue
Chesnay Schepler created FLINK-11269:
Summary: Extend Download page to list optional components
Key: FLINK-11269
URL: https://issues.apache.org/jira/browse/FLINK-11269
Project: Flink
Issu
Chesnay Schepler created FLINK-11266:
Summary: Only release hadoop-free Flink
Key: FLINK-11266
URL: https://issues.apache.org/jira/browse/FLINK-11266
Project: Flink
Issue Type: Improvemen
Chesnay Schepler created FLINK-11268:
Summary: Update deploy_staging_jars.sh to deploy multiple
flink-shaded-hadoop2 versions
Key: FLINK-11268
URL: https://issues.apache.org/jira/browse/FLINK-11268
Chesnay Schepler created FLINK-11267:
Summary: Update create_binary_release.sh to not create
hadoop-specific releases
Key: FLINK-11267
URL: https://issues.apache.org/jira/browse/FLINK-11267
Projec
Fokko Driesprong created FLINK-11265:
Summary: Invalid reference to AvroSinkWriter in example
AvroKeyValueSinkWriter
Key: FLINK-11265
URL: https://issues.apache.org/jira/browse/FLINK-11265
Project
Could it also work so that after so many tries it blacklists everything?
That way it would pretty much trigger a fresh restart.
Gyula
On Fri, 4 Jan 2019 at 10:11, Piotr Nowojski wrote:
> That’s a good point Till. Blacklisting TMs could be able to handle this.
> One scenario that might be proble
That’s a good point Till. Blacklisting TMs could be able to handle this. One
scenario that might be problematic is if clean restart is needed after a more
or less random number of job resubmissions, like if resource leakage has
different rates on different nodes. In such situation, if we blackli
Hi,
In that case I think instead of fixing the current back pressure monitoring
mechanism, it would be better to replace it with a new one based on output
queues length. But I haven’t thought it through, especially with respect to
performance implications, however my gut feeling is that it shou
10 matches
Mail list logo