[VOTE] Release 1.19.3, release candidate #1

[Ferenc Csaky <fe...@pm.me.INVALID>]

Hi everyone,

Please review and vote on the release candidate #1 for the version 1.19.3, 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 16AE0DDBBB2F380B [3],
* all artifacts to be deployed to the Maven Central Repository [4],
* source code tag "release-1.19.3-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,
Ferenc

[1] https://issues.apache.org/jira/secure/ReleaseNote.jspa?projectId=12315522&version=12355441
[2] https://dist.apache.org/repos/dist/dev/flink/flink-1.19.3-rc1/
[3] https://dist.apache.org/repos/dist/release/flink/KEYS
[4] https://repository.apache.org/content/repositories/orgapacheflink-1826/
[5] https://github.com/apache/flink/releases/tag/release-1.19.3-rc1
[6] https://github.com/apache/flink-web/pull/796

Re: [VOTE] Release 1.19.3, release candidate #1

[Ferenc Csaky <fe...@pm.me.INVALID>]

Since the tar archives are corrupt because of the usage of bsdtar, I close this thread
and create a new RC and will post a new vote thread for that.

Thanks,
Ferenc



On Tuesday, June 10th, 2025 at 17:50, Ferenc Csaky <fe...@pm.me.INVALID> wrote:

> 
> 
> I think the same issue just came up in a recent Flink Kubernetes Operator
> release let me take a look at that, and if necessary I will create a new RC.
> 
> Thanks for taking the time and figuring this out!
> 
> Best,
> Ferenc
> 
> 
> 
> 
> On Tuesday, June 10th, 2025 at 10:36, Sergey Nuyanzin snuyanzin@gmail.com wrote:
> 
> > thank you for the highlighting Ferenc!
> > 
> > I confirm that CI (including nightlies passed) for 1.19 which is great.
> > 
> > I started to check
> > - verified keys and signatures (ok)
> > - checked hashsums (ok)
> > - built from sources in Mac (ok)
> > - ran simple jobs (ok)
> > 
> > the thing which is not ok: I tried to build in Linux and it failed
> > with a number of files without license
> > after digging here it looks like if I decompress it with tar it
> > results to flink sources where for every file
> > there is a binary file with a almost same name (having prefix "._")
> > cat shows that every files starts with "Mac OS X"
> > I tend to think the reason is that probably for archives there were
> > used some Mac specific tools which lead to problems in Linux
> > 
> > On Sun, Jun 8, 2025 at 9:16 AM Ferenc Csaky ferenc.csaky@pm.me.invalid wrote:
> > 
> > > I would like to highlight that, now the Azure pipeline is fixed,
> > > there was a green CI run for the release-1.19 branch [1].
> > > 
> > > Best,
> > > Ferenc
> > > 
> > > [1] https://dev.azure.com/apache-flink/apache-flink/_build/results?buildId=68012&view=results
> > > 
> > > On Monday, June 2nd, 2025 at 17:36, Ferenc Csaky ferenc.csaky@pm.me.INVALID wrote:
> > > 
> > > > I do not necessarily see the direct connection between the CI
> > > > trigger and a new RC. Even if the CI trigger has to be fixed with
> > > > another commit into the `apache/flink` repo, my educated guess
> > > > would be it will not change anything in the product, so creating
> > > > a new RC would mean the exact same content.
> > > > 
> > > > Your point regarding we need to make sure we did not break anything
> > > > is fair, although I would like note that the ustream Nighly CI run [1]
> > > > and the one I ran on my fork with the rc1 tag [2] to produce wheels was
> > > > both successful, and AFAIK that covers pretty much everything that the
> > > > Azure CI run. I am not saying we should move forward, the GH CI is still
> > > > in beta, so we should definitely fix the Azure CI and get a green run,
> > > > but we should not drop RC1 yet.
> > > > 
> > > > Best,
> > > > Ferenc
> > > > 
> > > > [1] https://github.com/apache/flink/actions/runs/15382501322
> > > > [2] https://github.com/ferenc-csaky/flink/actions/runs/15333746515
> > > > 
> > > > On Monday, June 2nd, 2025 at 09:42, Sergey Nuyanzin snuyanzin@gmail.com wrote:
> > > > 
> > > > > Thanks for looking into this Ferenc
> > > > > 
> > > > > It seems I have to vote with
> > > > > -1 (binding)
> > > > > 
> > > > > I went through CI build and it looks like CI is broken for all
> > > > > branches (1.19.x, 1.20.x, 2.0.x it is run only for same old commit
> > > > > without taking others into account)
> > > > > It means no way to see whether there is anything else is broken or not
> > > > > with commits after that
> > > > > the issue for that https://issues.apache.org/jira/browse/FLINK-37883
> > 
> > --
> > Best regards,
> > Sergey

Re: [VOTE] Release 1.19.3, release candidate #1

[Ferenc Csaky <fe...@pm.me.INVALID>]

I think the same issue just came up in a recent Flink Kubernetes Operator
release let me take a look at that, and if necessary I will create a new RC.

Thanks for taking the time and figuring this out!

Best,
Ferenc




On Tuesday, June 10th, 2025 at 10:36, Sergey Nuyanzin <sn...@gmail.com> wrote:

> 
> 
> thank you for the highlighting Ferenc!
> 
> I confirm that CI (including nightlies passed) for 1.19 which is great.
> 
> I started to check
> - verified keys and signatures (ok)
> - checked hashsums (ok)
> - built from sources in Mac (ok)
> - ran simple jobs (ok)
> 
> the thing which is not ok: I tried to build in Linux and it failed
> with a number of files without license
> after digging here it looks like if I decompress it with tar it
> results to flink sources where for every file
> there is a binary file with a almost same name (having prefix "._")
> cat shows that every files starts with "Mac OS X"
> I tend to think the reason is that probably for archives there were
> used some Mac specific tools which lead to problems in Linux
> 
> On Sun, Jun 8, 2025 at 9:16 AM Ferenc Csaky ferenc.csaky@pm.me.invalid wrote:
> 
> > I would like to highlight that, now the Azure pipeline is fixed,
> > there was a green CI run for the release-1.19 branch [1].
> > 
> > Best,
> > Ferenc
> > 
> > [1] https://dev.azure.com/apache-flink/apache-flink/_build/results?buildId=68012&view=results
> > 
> > On Monday, June 2nd, 2025 at 17:36, Ferenc Csaky ferenc.csaky@pm.me.INVALID wrote:
> > 
> > > I do not necessarily see the direct connection between the CI
> > > trigger and a new RC. Even if the CI trigger has to be fixed with
> > > another commit into the `apache/flink` repo, my educated guess
> > > would be it will not change anything in the product, so creating
> > > a new RC would mean the exact same content.
> > > 
> > > Your point regarding we need to make sure we did not break anything
> > > is fair, although I would like note that the ustream Nighly CI run [1]
> > > and the one I ran on my fork with the rc1 tag [2] to produce wheels was
> > > both successful, and AFAIK that covers pretty much everything that the
> > > Azure CI run. I am not saying we should move forward, the GH CI is still
> > > in beta, so we should definitely fix the Azure CI and get a green run,
> > > but we should not drop RC1 yet.
> > > 
> > > Best,
> > > Ferenc
> > > 
> > > [1] https://github.com/apache/flink/actions/runs/15382501322
> > > [2] https://github.com/ferenc-csaky/flink/actions/runs/15333746515
> > > 
> > > On Monday, June 2nd, 2025 at 09:42, Sergey Nuyanzin snuyanzin@gmail.com wrote:
> > > 
> > > > Thanks for looking into this Ferenc
> > > > 
> > > > It seems I have to vote with
> > > > -1 (binding)
> > > > 
> > > > I went through CI build and it looks like CI is broken for all
> > > > branches (1.19.x, 1.20.x, 2.0.x it is run only for same old commit
> > > > without taking others into account)
> > > > It means no way to see whether there is anything else is broken or not
> > > > with commits after that
> > > > the issue for that https://issues.apache.org/jira/browse/FLINK-37883
> 
> 
> 
> 
> --
> Best regards,
> Sergey

Re: [VOTE] Release 1.19.3, release candidate #1

[Sergey Nuyanzin <sn...@gmail.com>]

thank you for the highlighting Ferenc!

I confirm that CI (including nightlies passed) for 1.19 which is great.

I started to check
- verified keys and signatures (ok)
- checked hashsums (ok)
- built from sources in Mac (ok)
- ran simple jobs (ok)

the thing which is not ok: I tried to build in Linux and it failed
with a number of files without license
after digging here it looks like if I decompress it with tar it
results to flink sources where for every file
there is a binary file with a almost same name (having prefix "._")
cat shows that every files starts with "Mac OS X"
I tend to think the reason is that probably for archives there were
used some Mac specific tools which lead to problems in Linux

On Sun, Jun 8, 2025 at 9:16 AM Ferenc Csaky <fe...@pm.me.invalid> wrote:
>
> I would like to highlight that, now the Azure pipeline is fixed,
> there was a green CI run for the release-1.19 branch [1].
>
> Best,
> Ferenc
>
> [1] https://dev.azure.com/apache-flink/apache-flink/_build/results?buildId=68012&view=results
>
>
>
> On Monday, June 2nd, 2025 at 17:36, Ferenc Csaky <fe...@pm.me.INVALID> wrote:
>
> >
> >
> > I do not necessarily see the direct connection between the CI
> > trigger and a new RC. Even if the CI trigger has to be fixed with
> > another commit into the `apache/flink` repo, my educated guess
> > would be it will not change anything in the product, so creating
> > a new RC would mean the exact same content.
> >
> > Your point regarding we need to make sure we did not break anything
> > is fair, although I would like note that the ustream Nighly CI run [1]
> > and the one I ran on my fork with the rc1 tag [2] to produce wheels was
> > both successful, and AFAIK that covers pretty much everything that the
> > Azure CI run. I am not saying we should move forward, the GH CI is still
> > in beta, so we should definitely fix the Azure CI and get a green run,
> > but we should not drop RC1 yet.
> >
> > Best,
> > Ferenc
> >
> > [1] https://github.com/apache/flink/actions/runs/15382501322
> > [2] https://github.com/ferenc-csaky/flink/actions/runs/15333746515
> >
> >
> > On Monday, June 2nd, 2025 at 09:42, Sergey Nuyanzin snuyanzin@gmail.com wrote:
> >
> > > Thanks for looking into this Ferenc
> > >
> > > It seems I have to vote with
> > > -1 (binding)
> > >
> > > I went through CI build and it looks like CI is broken for all
> > > branches (1.19.x, 1.20.x, 2.0.x it is run only for same old commit
> > > without taking others into account)
> > > It means no way to see whether there is anything else is broken or not
> > > with commits after that
> > > the issue for that https://issues.apache.org/jira/browse/FLINK-37883



-- 
Best regards,
Sergey

Re: [VOTE] Release 1.19.3, release candidate #1

[Ferenc Csaky <fe...@pm.me.INVALID>]

I would like to highlight that, now the Azure pipeline is fixed,
there was a green CI run for the release-1.19 branch [1].

Best,
Ferenc

[1] https://dev.azure.com/apache-flink/apache-flink/_build/results?buildId=68012&view=results



On Monday, June 2nd, 2025 at 17:36, Ferenc Csaky <fe...@pm.me.INVALID> wrote:

> 
> 
> I do not necessarily see the direct connection between the CI
> trigger and a new RC. Even if the CI trigger has to be fixed with
> another commit into the `apache/flink` repo, my educated guess
> would be it will not change anything in the product, so creating
> a new RC would mean the exact same content.
> 
> Your point regarding we need to make sure we did not break anything
> is fair, although I would like note that the ustream Nighly CI run [1]
> and the one I ran on my fork with the rc1 tag [2] to produce wheels was
> both successful, and AFAIK that covers pretty much everything that the
> Azure CI run. I am not saying we should move forward, the GH CI is still
> in beta, so we should definitely fix the Azure CI and get a green run,
> but we should not drop RC1 yet.
> 
> Best,
> Ferenc
> 
> [1] https://github.com/apache/flink/actions/runs/15382501322
> [2] https://github.com/ferenc-csaky/flink/actions/runs/15333746515
> 
> 
> On Monday, June 2nd, 2025 at 09:42, Sergey Nuyanzin snuyanzin@gmail.com wrote:
> 
> > Thanks for looking into this Ferenc
> > 
> > It seems I have to vote with
> > -1 (binding)
> > 
> > I went through CI build and it looks like CI is broken for all
> > branches (1.19.x, 1.20.x, 2.0.x it is run only for same old commit
> > without taking others into account)
> > It means no way to see whether there is anything else is broken or not
> > with commits after that
> > the issue for that https://issues.apache.org/jira/browse/FLINK-37883

Re: [VOTE] Release 1.19.3, release candidate #1

[Ferenc Csaky <fe...@pm.me.INVALID>]

I do not necessarily see the direct connection between the CI
trigger and a new RC. Even if the CI trigger has to be fixed with
another commit into the `apache/flink` repo, my educated guess
would be it will not change anything in the product, so creating
a new RC would mean the exact same content.

Your point regarding we need to make sure we did not break anything
is fair, although I would like note that the ustream Nighly CI run [1]
and the one I ran on my fork with the rc1 tag [2] to produce wheels was
both successful, and AFAIK that covers pretty much everything that the
Azure CI run. I am not saying we should move forward, the GH CI is still
in beta, so we should definitely fix the Azure CI and get a green run,
but we should not drop RC1 yet.

Best,
Ferenc

[1] https://github.com/apache/flink/actions/runs/15382501322
[2] https://github.com/ferenc-csaky/flink/actions/runs/15333746515


On Monday, June 2nd, 2025 at 09:42, Sergey Nuyanzin <sn...@gmail.com> wrote:

> 
> 
> Thanks for looking into this Ferenc
> 
> It seems I have to vote with
> -1 (binding)
> 
> I went through CI build and it looks like CI is broken for all
> branches (1.19.x, 1.20.x, 2.0.x it is run only for same old commit
> without taking others into account)
> It means no way to see whether there is anything else is broken or not
> with commits after that
> the issue for that https://issues.apache.org/jira/browse/FLINK-37883

Re: [VOTE] Release 1.19.3, release candidate #1

[Sergey Nuyanzin <sn...@gmail.com>]

Thanks for looking into this Ferenc

It seems I have to vote with
-1 (binding)

I went through CI build and it looks like CI is broken for all
branches (1.19.x, 1.20.x, 2.0.x it is run only for same old commit
without taking others into account)
It means no way to see whether there is anything else is broken or not
with commits after that
the issue for that https://issues.apache.org/jira/browse/FLINK-37883

Re: [VOTE] Release 1.19.3, release candidate #1

[Ferenc Csaky <fe...@pm.me.INVALID>]

I also experienced the MacOS fail because of fastavro, but that should be fixed by
FLINK-37804 that is backported to all currently active release branches [1]. No
Azure CI run picked this latest state yet.

FYI, since we talk about wheel files already, the wheel files for the release now
built via the GH CI Nightly, not the Azure pipeline, which was done in FLINK-34487
and also backported to all active release branches.

During this release I also updated the relevant parts of the release guide doc [2].

Best,
Ferenc

[1] https://github.com/apache/flink/commit/858008f6fd1ca7c29d2970da09b24600288b3e73
[2] https://cwiki.apache.org/confluence/pages/viewpage.action?pageId=73631092#CreatingaFlinkRelease-BuildandstageJavaandPythonartifacts


On Friday, May 30th, 2025 at 20:32, Sergey Nuyanzin <sn...@gmail.com> wrote:

> 
> 
> to be more precise: last time "build_wheels_on_macos" was green is 30.04.2025
> after that every time it fail
> I created jira issue for that https://issues.apache.org/jira/browse/FLINK-37880
> 
> On Fri, May 30, 2025 at 8:03 PM Sergey Nuyanzin snuyanzin@gmail.com wrote:
> 
> > Do we know why all ci [1] are red since 28.05.2025?
> > 
> > Not sure if it might block the release...
> > 
> > [1] https://dev.azure.com/apache-flink/apache-flink/_build?definitionId=1
> > 
> > Best regards,
> > Sergey
> > 
> > On Fri, May 30, 2025, 13:29 Ferenc Csaky ferenc.csaky@pm.me.invalid wrote:
> > 
> > > Hi everyone,
> > > 
> > > Please review and vote on the release candidate #1 for the version 1.19.3, 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 16AE0DDBBB2F380B [3],
> > > * all artifacts to be deployed to the Maven Central Repository [4],
> > > * source code tag "release-1.19.3-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,
> > > Ferenc
> > > 
> > > [1] https://issues.apache.org/jira/secure/ReleaseNote.jspa?projectId=12315522&version=12355441
> > > [2] https://dist.apache.org/repos/dist/dev/flink/flink-1.19.3-rc1/
> > > [3] https://dist.apache.org/repos/dist/release/flink/KEYS
> > > [4] https://repository.apache.org/content/repositories/orgapacheflink-1826/
> > > [5] https://github.com/apache/flink/releases/tag/release-1.19.3-rc1
> > > [6] https://github.com/apache/flink-web/pull/796
> 
> 
> 
> 
> --
> Best regards,
> Sergey

Re: [VOTE] Release 1.19.3, release candidate #1

[Sergey Nuyanzin <sn...@gmail.com>]

to be more precise: last time "build_wheels_on_macos" was green is 30.04.2025
after that every time it fail
I created jira issue for that https://issues.apache.org/jira/browse/FLINK-37880

On Fri, May 30, 2025 at 8:03 PM Sergey Nuyanzin <sn...@gmail.com> wrote:
>
> Do we know why all ci [1] are red since 28.05.2025?
>
> Not sure if it might block the release...
>
> [1] https://dev.azure.com/apache-flink/apache-flink/_build?definitionId=1
>
> Best regards,
> Sergey
>
> On Fri, May 30, 2025, 13:29 Ferenc Csaky <fe...@pm.me.invalid> wrote:
>>
>> Hi everyone,
>>
>> Please review and vote on the release candidate #1 for the version 1.19.3, 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 16AE0DDBBB2F380B [3],
>> * all artifacts to be deployed to the Maven Central Repository [4],
>> * source code tag "release-1.19.3-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,
>> Ferenc
>>
>> [1] https://issues.apache.org/jira/secure/ReleaseNote.jspa?projectId=12315522&version=12355441
>> [2] https://dist.apache.org/repos/dist/dev/flink/flink-1.19.3-rc1/
>> [3] https://dist.apache.org/repos/dist/release/flink/KEYS
>> [4] https://repository.apache.org/content/repositories/orgapacheflink-1826/
>> [5] https://github.com/apache/flink/releases/tag/release-1.19.3-rc1
>> [6] https://github.com/apache/flink-web/pull/796



-- 
Best regards,
Sergey

Re: [VOTE] Release 1.19.3, release candidate #1

[Sergey Nuyanzin <sn...@gmail.com>]

Do we know why all ci [1] are red since 28.05.2025?

Not sure if it might block the release...

[1] https://dev.azure.com/apache-flink/apache-flink/_build?definitionId=1

Best regards,
Sergey

On Fri, May 30, 2025, 13:29 Ferenc Csaky <fe...@pm.me.invalid> wrote:

> Hi everyone,
>
> Please review and vote on the release candidate #1 for the version 1.19.3,
> 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
> 16AE0DDBBB2F380B [3],
> * all artifacts to be deployed to the Maven Central Repository [4],
> * source code tag "release-1.19.3-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,
> Ferenc
>
> [1]
> https://issues.apache.org/jira/secure/ReleaseNote.jspa?projectId=12315522&version=12355441
> [2] https://dist.apache.org/repos/dist/dev/flink/flink-1.19.3-rc1/
> [3] https://dist.apache.org/repos/dist/release/flink/KEYS
> [4]
> https://repository.apache.org/content/repositories/orgapacheflink-1826/
> [5] https://github.com/apache/flink/releases/tag/release-1.19.3-rc1
> [6] https://github.com/apache/flink-web/pull/796
>