Hi Aljoscha,
That'll be great. Look forward to a better Kinesis connector from
Flink!
Thanks,
Bowen
On Fri, Aug 4, 2017 at 11:07 PM, Aljoscha Krettek
wrote:
> The voting time has passed and I'm happy to announce that we've collected
> enough votes to release this RC as Flink 1.3.2.
>
> +1
The voting time has passed and I'm happy to announce that we've collected
enough votes to release this RC as Flink 1.3.2.
+1 votes:
- Stefan (non-binding)
- Till (binding)
- Gordon (binding)
- Piotr (non-binding)
- Fabian (binding)
- Aljoscha (binding)
- Greg (binding)
That's 7 votes, 5 binding.
With 24 hours now having passed, this vote is now complete. I'll summarise the
results in an extra thread.
Best,
Aljoscha
> On 5. Aug 2017, at 07:49, Aljoscha Krettek wrote:
>
> Hi Bowen,
>
> You're right, I didn't give a lot of thought to Kinesis because there were
> critical bugs in Kafka a
Hi Bowen,
You're right, I didn't give a lot of thought to Kinesis because there were
critical bugs in Kafka and incremental checkpointing that needed to be fixed
quickly. It's not good to neglect another important connector for this, though.
We should start focusing on the Kinesis issues next w
Thanks Aljoscha!
+1 (binding)
- verified source and binary signatures
- verified source and binary checksums
- verified LICENSEs
- verified NOTICEs
- built from source
Greg
> On Aug 4, 2017, at 2:00 AM, Aljoscha Krettek wrote:
>
> Hi everyone,
>
> Please review and vote on the release candid
Hi everyone,
I read the doc, and found that we are not even testing Flink against
Kinesis with *flink-connector-kinesis*.
According to our experience, *flink-connector-kinesis *releases tend to
have more issues w.r.t. dependency version conflicts, performance,
capability, and stability. We've rep
+1
- checked checksums and signatures
- did extensive cluster testing on GCE with dataproc and a lot of different
combinations of backends and configuration settings
Tested configuration combinations (from the testing coordination document):
- (aljoscha, GCE, dataproc, Rocks(incremental), YAR
+1
- checked added and modified Maven dependencies since Flink 1.3.1
- io.dropwizard.metrics/metrics-core was upgraded to 3.1.5 and is ASL
- org.xerial.snappy/snappy-java/1.0.5 was added and is ASL
- checked flink-runtime-web for new files with non-Apache licenses
Cheers, Fabian
2017-08-04 1
+1
- Sources contain no binaries
- Build and test (men clean verify) from source archive, found some issues
previously mentioned by Nico
- Build with Scala 2.10/Scala 2.11
- Test start up scripts for local and standalone mode
Cluster tests:
- verified that FlinkKafkaProducer010 provides at-leas
+1
- Functional cluster test results carried over from RC2 (did not retest those
for RC3 as the additional commit fix for Gelly should be irrelevant)
- Built from source with Scala 2.10 and 2.11 (MacOSX), successful
- Checked to-be-released Maven artifacts
Cheers,
Gordon
On 4 August 2017 at 5:4
+1
- Checked checksums and GPG files
- Sources contain no binaries
- Check all POM files point to the same version
- Read readme file
- Build from source archive with Hadoop 2.8.1
- Test start up scripts for local and standalone mode
- Test job submission for multi-node cluster
- Plan-visualizer
-
+1
1. Cluster tests
Tested with the stateful state machine job on the following settings:
- Cloud env: AWS
- Distributions: EMR 5.7.0
- Flink deployment method: YARN (Hadoop 2.7.2)
- HA: enabled
- Kerberos: disabled
- Kafka version: 0.10, 0.11
- State Backends: Heap (Sync / Async) & RocksDB (incr
Hi everyone,
Please review and vote on the release candidate #3 for the version 1.3.2, 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],
13 matches
Mail list logo
