I've filed https://issues.apache.org/jira/browse/FLINK-20519 for the hbase issue.

Since we still ship the protobuf license I don't think this is a blocker.

On 12/7/2020 10:52 AM, Dawid Wysakowicz wrote:

+1 (binding)

* Verified the checksums
* Verified that the source archives do not contain any binaries
* Built Flink from sources
* Run a streaming WorldCount example in BATCH and STREAM mode
* Run a slightly heavier WorldCount version in BATCH and STREAM mode
* Verified licensing of Hbase connectors
*** I found one issue that we do not list protobuf-java-util in flink-sql-connector-hbase-2.2 NOTICE file, which is pulled in through hbase-shaded-miscellaneous (there is a couple more dependencies we do not list from it, but they are Apache licensed)*
* Run some Table examples (I found two examples are not runnable[1]), but I 
would not consider it a blocker

Let me know what you think about the licensing issue.

Best,
Dawid

[1]https://issues.apache.org/jira/browse/FLINK-20464
On 07/12/2020 08:07, Leonard Xu wrote:
+1 (non-binding)

- checked/verified signatures and hashes
- built from source code with scala 2.11 succeeded
- checked that there are no missing artifacts
- started a cluster, WebUI was accessible, submitted a wordcount job and ran 
succeeded, no suspicious log output
- tested using SQL Client to submit job and the query result is expected
- tested read/write from/to sql kafka/upsert-kafka connector in SQL Client
- tested read/write/join hvie table in SQL Client

Best,
Leonard Xu


在 2020年12月7日,11:17,Guowei Ma<guowei....@gmail.com>  写道:

+1(non-binding)
- build from source
- build a docker image
- start a session from local k8s cluster
- submit a wordcount job in streaming mode.
- submit a wordcount job in batch mode.
Best,
Guowei


On Sat, Dec 5, 2020 at 3:13 PM Zhu Zhu<reed...@gmail.com>  wrote:

+1 (binding)

- verified signature and checksum
- built from source
- run testing jobs on yarn with manually triggered failures. checked logs
and WebUI of those jobs
  * DataStream job (paralelism=1000) with multiple disjoint pipelined
regions
  * DataSet job (paralelism=1000) with all edges blocking

Thanks,
Zhu

Till Rohrmann<trohrm...@apache.org>  于2020年12月4日周五 下午11:45写道:

+1 (binding)

* Verified the checksums
* Ran RC on Minikube cluster
** Session mode
** Application mode
* Built Flink from sources

Cheers,
Till

On Fri, Dec 4, 2020 at 2:15 PM Wei Zhong<weizhong0...@gmail.com>  wrote:

+1 (non-binding)

- verified checksums and signatures
- build Flink with Scala 2.11
- pip install pyflink on Windows python 3.7
- run a python job with udfs on Windows
- pyflink shell works well on local mode and remote mode

Best,
Wei

在 2020年12月4日,17:21,Yang Wang<danrtsey...@gmail.com>  写道:

+1 (non-binding)

* Build from source
* Deploy Flink cluster in following deployments with HA
enabled(ZooKeeper
and K8s), including kill JobManager and check failover
* Native K8s Session
* Native K8s Application
* Yarn Session
* Yarn Per-Job
* Yarn Application
* Check webUI and logs in different deployments especially via
`kubectl
logs` in K8s

Best,
Yang

Xintong Song<tonysong...@gmail.com>  于2020年12月4日周五 下午3:00写道:

+1 (non-binding)

  - Verified checksums and signatures
  - No binaries found in source archive
  - Build from source
  - Tried a couple of example jobs in various deployment mode
     - Local
     - Standalone
     - Native Kubernetes Application
     - Native Kubernetes Session
     - Yarn Job
     - Yarn Session
  - Changing memory configurations, things work as expected
  - UI looks good
  - Logs look good



Thank you~

Xintong Song



On Thu, Dec 3, 2020 at 9:18 PM Rui Li<lirui.fu...@gmail.com>
wrote:
+1 (non-binding)

Built from source and verified hive connector tests for different
hive
versions.
Setup a cluster to connect to a real hive warehouse and run some
queries
successfully.

On Thu, Dec 3, 2020 at 8:44 PM Xingbo Huang<hxbks...@gmail.com>
wrote:
+1 (non-binding)

Checks:
1. verified checksums and signatures
2. build Flink with Scala 2.11
3. pip install pyflink in MacOS/CentOS under py35,py36,py37,py38
4. test Pandas UDAF/General UDAF/Python DataStream MapFunction
5. start standalone cluster and submit a python udf job.
6. verified NOTICE/LICENSE files of some regular modules

I observed that the NOTICE file of flink-sql-connector-hbase-2.2
lists
3
dependencies that are not bundled in:
commons-lang:commons-lang:2.6
org.apache.hbase:hbase-hadoop-compat:2.2.3
org.apache.hbase:hbase-hadoop2-compat:2.2.3

I guess listing more than dependencies with apache licensed
shouldn't
be
a
blocker issue. I have opened a PR[1] to fix it.

[1]https://github.com/apache/flink/pull/14299

Best,
Xingbo

Robert Metzger<rmetz...@apache.org>  于2020年12月3日周四 下午5:36写道:

There's now a pull request for the announcement blog post, please
help
checking it:https://github.com/apache/flink-web/pull/397

On Thu, Dec 3, 2020 at 9:03 AM Robert Metzger <
rmetz...@apache.org
wrote:
+1 (binding)


Checks:
- checksums seem correct
- source archive code compiles
- Compiled a test job against the staging repository
- launched a standalone cluster, ran some test jobs against it
- quickstart contains correct version
- regular jars contain correct NOTICE file
- Looked a bit over the output of
    git diff release-1.11.2...release-1.12 --  "**/pom.xml"



I noticed that at least one more jar file contains an invalid
LICENSE
file
in it's root. This has already been the case with Flink 1.11,
and
from
the
context (apache flink jar, all the other license and notice
files
talk
about this being an Apache project) it should be clear that the
license
file is not meant for the whole jar file contents.
I will still extend the automated LicenseChecker to resolve
this,
but I
don't want to cancel the release because of this.



On Wed, Dec 2, 2020 at 11:19 AM Robert Metzger <
rmetz...@apache.org>
wrote:

Hi everyone,

We have resolved the licensing issue Chesnay found.

Please review and vote on the release candidate #3 for the
version
1.12.0, 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 [1a], and website release notes [1b]
* 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 D9839159 [3],
* all artifacts to be deployed to the Maven Central Repository
[4],
* source code tag "release-1.12.0-rc3" [5]

We will soon publish the PR for the release announcement blog
post!
The vote will be open for at least 72 hours. It is adopted by
majority
approval, with at least 3 PMC affirmative votes.

Thanks,
Dian & Robert

[1a]

https://issues.apache.org/jira/secure/ReleaseNote.jspa?projectId=12315522&version=12348263
[1b]https://github.com/apache/flink/pull/14195
[2]
https://dist.apache.org/repos/dist/dev/flink/flink-1.12.0-rc3/
[3]https://dist.apache.org/repos/dist/release/flink/KEYS
[4]

https://repository.apache.org/content/repositories/orgapacheflink-1404
[5]
https://github.com/apache/flink/releases/tag/release-1.12.0-rc3
--
Best regards!
Rui Li


Reply via email to