our feedback,
suggestions, and any potential concerns you might have.
Thanks,
Zexian Wu
+1 (non-binding)
I checked:
- Verified source archive signature with key 125FD8DB (Danny Cranmer).
- SHA512 checksum matched between generated and official .sha512 file.
- Built source successfully with JDK8 & Maven 3.8.6.
- Confirmed all poms use version 5.1.0-1.20.
Best regards,
Wu Zexian
+1 (non-binding)
I checked:
- Verified source archive signature with key 125FD8DB (Danny Cranmer).
- SHA512 checksum matched between generated and official .sha512 file.
- Built source successfully with JDK8 & Maven 3.8.6.
- Confirmed all poms use version 5.1.0-1.20.
Best regards,
Wu Zexian
+1 (non-binding)
Checklist:
- verified tarball checksums & gpg signatures are valid
- built from source successfully on macOS with JDK 17 and Maven 3.8.x
- started a local standalone cluster and verified the Web UI and logs
- ran simple queries via the sql-client in batch mode
- confirmed the git
ign integrating it into the source
> operator lifecycle, it is already able to meet the vast majority of
> scenarios, looks good from my side.
>
> Best,
> Leonard
>
>
> > 2025 7月 18 12:01，Zexian WU 写道：
> >
> > Hi everyone,
> > I would like to star
+1 (non-binding)
Checklist:
- verified tarball checksums & gpg signatures are valid
- built from source successfully on macOS with JDK 17 and Maven 3.8.x
- started a local standalone cluster and verified the Web UI and logs
- ran simple queries via the sql-client in batch mode
- confirmed the git
+1 (non-binding)
Checklist:
- verified tarball checksums & gpg signatures are valid
- built from source successfully on macOS with JDK 17 and Maven 3.8.x
- confirmed the git release tag `release-2.1.0-rc1` is correct
Best Regards,
Zexian
> 2025年7月21日 21:40，Ron Liu 写道：
>
> Hi everyone,
>
ign integrating it into the source
> operator lifecycle, it is already able to meet the vast majority of
> scenarios, looks good from my side.
>
> Best,
> Leonard
>
>
> > 2025 7月 18 12:01，Zexian WU 写道：
> >
> > Hi everyone,
> > I would like to star
it.record-per-second`. It seems only FLIP-27 source supports
> rate-limit, it's better we can throw an exception to notify users if the
> source doesn't support this feature.
>
> Best,
> Shengkai
>
> Zexian WU 于2025年7月28日周一 11:52写道：
>
> > Hi Leonard,
>
each operator's qps by
>ConfigOption. This is convenient if we can get the operatorId
> easily
>before submitting. But in open source, it may be hard to do so. I wonder
>whether there exist similar way so that we do not need to implement each
>connector.
>
>
operators. notifyCheckpointAborted is also not guaranteed to be called, so
>
> if there is a risk of deadlock, we should find other options.
>
> Best,
>
> Arvid
>
>
> On Tue, Aug 5, 2025 at 8:48
> AM Jiangang Liu <
> liujiangangp...@gmail.com>
> wrote:
>
> &
your explanation.
> > 1. It makes sense to me to solve the imbalance in other issues.
> > 2. In fact, we support to adjust the source's qps dynamically in our
> inner
> > flink. We can support this feature later.
> >
> > Best,
> > Jiangang Liu
> >
&
12 matches
Mail list logo
