You are viewing a plain text version of this content. The canonical link for it is here.
Posted to dev@nifi.apache.org by Marton Szasz <sz...@apache.org> on 2021/01/21 14:55:11 UTC
[DISCUSS] MiNiFi C++ 0.9.0 release
Hi community,
I'd like to initiate a discussion about the next release of MiNiFi C++. The
last release was over a year ago, and there have been countless new
features, bug fixes and stability improvements committed to the development
branch since then.
Some highlights from the last year:
- Added support for RocksDB-based content repository for better performance
- Added SQL extension
- Improved task scheduling
- Various C2 improvements
- Bug fixes and improvements to TailFile, ConsumeWindowsEventLog,
MergeContent, CompressContent, PublishKafka, InvokeHTTP
- Implemented RetryFlowFile and smart handling of loopback connections
- Added a way to encrypt sensitive config properties and the flow
configuration
- Implemented full S3 support
- Reduced memory footprint when working with many flow files
Why 0.9.0 instead of 0.8.0?
The codebase is starting to stabilize and we're getting closer to a 1.0
milestone, but I feel like the extension API is not yet something we can
commit to support in the long term. Additionally, a release branch for
0.8.0 has been created earlier, but it was not released, and with all the
new development it wouldn't make sense to release it.
Do you think 0.9.0 is a good target, or would you go with 0.8.0 or 1.0
instead? Are there any blockers that should definitely make it into the
next release?
Thanks,
Marton
Re: [DISCUSS] MiNiFi C++ 0.9.0 release
Posted by Marton Szasz <sz...@gmail.com>.
We found some release blockers, so it may take a bit longer for me to
prepare the RC.
- The license shown during installation on windows (msi/LICENSE.txt) was
not kept in sync with LICENSE. This is waiting for review.
- New tests in a pull request uncovered an issue with delete_dir that
recursively removed all files from the drive if called without arguments.
We're not aware of any actual calls without an argument, but Gabor fixed it
to be safe. This is waiting for review.
- A crash on startup in Windows debug builds was reported by Arpad. This is
under investigation.
The release is tracked by MINIFICPP-1463
<https://issues.apache.org/jira/browse/MINIFICPP-1463>. I'm hoping to
prepare an RC by the end of this week or early next week.
On Tue, 26 Jan 2021 at 15:56, Pierre Villard <pi...@gmail.com>
wrote:
> Thanks Marton, I think it all makes sense!
>
> Le mar. 26 janv. 2021 à 18:54, Marton Szasz <sz...@apache.org> a écrit :
>
> > I moved the issues to 0.9.0 and Arpad removed 0.8.0 from Jira.
> >
> > Judging from the lack of replies, there seem to be no blocking issues.
> > I volunteer to take RM duties for the release.
> >
> > Marton
> >
> > On Tue, 26 Jan 2021 at 15:40, Arpad Boda <ab...@apache.org> wrote:
> >
> > > I think moving the tickets to 0.9.0 is easier, so +1 for 0.9.0!
> > >
> > > On Thu, Jan 21, 2021 at 4:17 PM Joe Witt <jo...@gmail.com> wrote:
> > >
> > > > Marton,
> > > >
> > > > I think if consensus matches your view on api stability then 0.9.0 is
> > > most
> > > > appropriate over a 1.0 for now. The progress made here in the past
> > year
> > > is
> > > > really awesome and frankly knowing some of the uses of this at the
> > scale
> > > > and diversity of usage is really exciting.
> > > >
> > > > You would just remove the 0.8.0 JIRA target and ensure any tickets
> > tagged
> > > > there are tagged to 0.9.0? Or just go back to 0.8.0 and get rid of
> > 0.9?
> > > >
> > > > Thanks
> > > >
> > > > On Thu, Jan 21, 2021 at 7:55 AM Marton Szasz <sz...@apache.org>
> > wrote:
> > > >
> > > > > Hi community,
> > > > >
> > > > > I'd like to initiate a discussion about the next release of MiNiFi
> > C++.
> > > > The
> > > > > last release was over a year ago, and there have been countless new
> > > > > features, bug fixes and stability improvements committed to the
> > > > development
> > > > > branch since then.
> > > > >
> > > > > Some highlights from the last year:
> > > > > - Added support for RocksDB-based content repository for better
> > > > performance
> > > > > - Added SQL extension
> > > > > - Improved task scheduling
> > > > > - Various C2 improvements
> > > > > - Bug fixes and improvements to TailFile, ConsumeWindowsEventLog,
> > > > > MergeContent, CompressContent, PublishKafka, InvokeHTTP
> > > > > - Implemented RetryFlowFile and smart handling of loopback
> > connections
> > > > > - Added a way to encrypt sensitive config properties and the flow
> > > > > configuration
> > > > > - Implemented full S3 support
> > > > > - Reduced memory footprint when working with many flow files
> > > > >
> > > > > Why 0.9.0 instead of 0.8.0?
> > > > > The codebase is starting to stabilize and we're getting closer to a
> > 1.0
> > > > > milestone, but I feel like the extension API is not yet something
> we
> > > can
> > > > > commit to support in the long term. Additionally, a release branch
> > for
> > > > > 0.8.0 has been created earlier, but it was not released, and with
> all
> > > the
> > > > > new development it wouldn't make sense to release it.
> > > > >
> > > > > Do you think 0.9.0 is a good target, or would you go with 0.8.0 or
> > 1.0
> > > > > instead? Are there any blockers that should definitely make it into
> > the
> > > > > next release?
> > > > >
> > > > > Thanks,
> > > > > Marton
> > > > >
> > > >
> > >
> >
>
Re: [DISCUSS] MiNiFi C++ 0.9.0 release
Posted by Pierre Villard <pi...@gmail.com>.
Thanks Marton, I think it all makes sense!
Le mar. 26 janv. 2021 à 18:54, Marton Szasz <sz...@apache.org> a écrit :
> I moved the issues to 0.9.0 and Arpad removed 0.8.0 from Jira.
>
> Judging from the lack of replies, there seem to be no blocking issues.
> I volunteer to take RM duties for the release.
>
> Marton
>
> On Tue, 26 Jan 2021 at 15:40, Arpad Boda <ab...@apache.org> wrote:
>
> > I think moving the tickets to 0.9.0 is easier, so +1 for 0.9.0!
> >
> > On Thu, Jan 21, 2021 at 4:17 PM Joe Witt <jo...@gmail.com> wrote:
> >
> > > Marton,
> > >
> > > I think if consensus matches your view on api stability then 0.9.0 is
> > most
> > > appropriate over a 1.0 for now. The progress made here in the past
> year
> > is
> > > really awesome and frankly knowing some of the uses of this at the
> scale
> > > and diversity of usage is really exciting.
> > >
> > > You would just remove the 0.8.0 JIRA target and ensure any tickets
> tagged
> > > there are tagged to 0.9.0? Or just go back to 0.8.0 and get rid of
> 0.9?
> > >
> > > Thanks
> > >
> > > On Thu, Jan 21, 2021 at 7:55 AM Marton Szasz <sz...@apache.org>
> wrote:
> > >
> > > > Hi community,
> > > >
> > > > I'd like to initiate a discussion about the next release of MiNiFi
> C++.
> > > The
> > > > last release was over a year ago, and there have been countless new
> > > > features, bug fixes and stability improvements committed to the
> > > development
> > > > branch since then.
> > > >
> > > > Some highlights from the last year:
> > > > - Added support for RocksDB-based content repository for better
> > > performance
> > > > - Added SQL extension
> > > > - Improved task scheduling
> > > > - Various C2 improvements
> > > > - Bug fixes and improvements to TailFile, ConsumeWindowsEventLog,
> > > > MergeContent, CompressContent, PublishKafka, InvokeHTTP
> > > > - Implemented RetryFlowFile and smart handling of loopback
> connections
> > > > - Added a way to encrypt sensitive config properties and the flow
> > > > configuration
> > > > - Implemented full S3 support
> > > > - Reduced memory footprint when working with many flow files
> > > >
> > > > Why 0.9.0 instead of 0.8.0?
> > > > The codebase is starting to stabilize and we're getting closer to a
> 1.0
> > > > milestone, but I feel like the extension API is not yet something we
> > can
> > > > commit to support in the long term. Additionally, a release branch
> for
> > > > 0.8.0 has been created earlier, but it was not released, and with all
> > the
> > > > new development it wouldn't make sense to release it.
> > > >
> > > > Do you think 0.9.0 is a good target, or would you go with 0.8.0 or
> 1.0
> > > > instead? Are there any blockers that should definitely make it into
> the
> > > > next release?
> > > >
> > > > Thanks,
> > > > Marton
> > > >
> > >
> >
>
Re: [DISCUSS] MiNiFi C++ 0.9.0 release
Posted by Marton Szasz <sz...@apache.org>.
I moved the issues to 0.9.0 and Arpad removed 0.8.0 from Jira.
Judging from the lack of replies, there seem to be no blocking issues.
I volunteer to take RM duties for the release.
Marton
On Tue, 26 Jan 2021 at 15:40, Arpad Boda <ab...@apache.org> wrote:
> I think moving the tickets to 0.9.0 is easier, so +1 for 0.9.0!
>
> On Thu, Jan 21, 2021 at 4:17 PM Joe Witt <jo...@gmail.com> wrote:
>
> > Marton,
> >
> > I think if consensus matches your view on api stability then 0.9.0 is
> most
> > appropriate over a 1.0 for now. The progress made here in the past year
> is
> > really awesome and frankly knowing some of the uses of this at the scale
> > and diversity of usage is really exciting.
> >
> > You would just remove the 0.8.0 JIRA target and ensure any tickets tagged
> > there are tagged to 0.9.0? Or just go back to 0.8.0 and get rid of 0.9?
> >
> > Thanks
> >
> > On Thu, Jan 21, 2021 at 7:55 AM Marton Szasz <sz...@apache.org> wrote:
> >
> > > Hi community,
> > >
> > > I'd like to initiate a discussion about the next release of MiNiFi C++.
> > The
> > > last release was over a year ago, and there have been countless new
> > > features, bug fixes and stability improvements committed to the
> > development
> > > branch since then.
> > >
> > > Some highlights from the last year:
> > > - Added support for RocksDB-based content repository for better
> > performance
> > > - Added SQL extension
> > > - Improved task scheduling
> > > - Various C2 improvements
> > > - Bug fixes and improvements to TailFile, ConsumeWindowsEventLog,
> > > MergeContent, CompressContent, PublishKafka, InvokeHTTP
> > > - Implemented RetryFlowFile and smart handling of loopback connections
> > > - Added a way to encrypt sensitive config properties and the flow
> > > configuration
> > > - Implemented full S3 support
> > > - Reduced memory footprint when working with many flow files
> > >
> > > Why 0.9.0 instead of 0.8.0?
> > > The codebase is starting to stabilize and we're getting closer to a 1.0
> > > milestone, but I feel like the extension API is not yet something we
> can
> > > commit to support in the long term. Additionally, a release branch for
> > > 0.8.0 has been created earlier, but it was not released, and with all
> the
> > > new development it wouldn't make sense to release it.
> > >
> > > Do you think 0.9.0 is a good target, or would you go with 0.8.0 or 1.0
> > > instead? Are there any blockers that should definitely make it into the
> > > next release?
> > >
> > > Thanks,
> > > Marton
> > >
> >
>
Re: [DISCUSS] MiNiFi C++ 0.9.0 release
Posted by Arpad Boda <ab...@apache.org>.
I think moving the tickets to 0.9.0 is easier, so +1 for 0.9.0!
On Thu, Jan 21, 2021 at 4:17 PM Joe Witt <jo...@gmail.com> wrote:
> Marton,
>
> I think if consensus matches your view on api stability then 0.9.0 is most
> appropriate over a 1.0 for now. The progress made here in the past year is
> really awesome and frankly knowing some of the uses of this at the scale
> and diversity of usage is really exciting.
>
> You would just remove the 0.8.0 JIRA target and ensure any tickets tagged
> there are tagged to 0.9.0? Or just go back to 0.8.0 and get rid of 0.9?
>
> Thanks
>
> On Thu, Jan 21, 2021 at 7:55 AM Marton Szasz <sz...@apache.org> wrote:
>
> > Hi community,
> >
> > I'd like to initiate a discussion about the next release of MiNiFi C++.
> The
> > last release was over a year ago, and there have been countless new
> > features, bug fixes and stability improvements committed to the
> development
> > branch since then.
> >
> > Some highlights from the last year:
> > - Added support for RocksDB-based content repository for better
> performance
> > - Added SQL extension
> > - Improved task scheduling
> > - Various C2 improvements
> > - Bug fixes and improvements to TailFile, ConsumeWindowsEventLog,
> > MergeContent, CompressContent, PublishKafka, InvokeHTTP
> > - Implemented RetryFlowFile and smart handling of loopback connections
> > - Added a way to encrypt sensitive config properties and the flow
> > configuration
> > - Implemented full S3 support
> > - Reduced memory footprint when working with many flow files
> >
> > Why 0.9.0 instead of 0.8.0?
> > The codebase is starting to stabilize and we're getting closer to a 1.0
> > milestone, but I feel like the extension API is not yet something we can
> > commit to support in the long term. Additionally, a release branch for
> > 0.8.0 has been created earlier, but it was not released, and with all the
> > new development it wouldn't make sense to release it.
> >
> > Do you think 0.9.0 is a good target, or would you go with 0.8.0 or 1.0
> > instead? Are there any blockers that should definitely make it into the
> > next release?
> >
> > Thanks,
> > Marton
> >
>
Re: [DISCUSS] MiNiFi C++ 0.9.0 release
Posted by Joe Witt <jo...@gmail.com>.
Marton,
I think if consensus matches your view on api stability then 0.9.0 is most
appropriate over a 1.0 for now. The progress made here in the past year is
really awesome and frankly knowing some of the uses of this at the scale
and diversity of usage is really exciting.
You would just remove the 0.8.0 JIRA target and ensure any tickets tagged
there are tagged to 0.9.0? Or just go back to 0.8.0 and get rid of 0.9?
Thanks
On Thu, Jan 21, 2021 at 7:55 AM Marton Szasz <sz...@apache.org> wrote:
> Hi community,
>
> I'd like to initiate a discussion about the next release of MiNiFi C++. The
> last release was over a year ago, and there have been countless new
> features, bug fixes and stability improvements committed to the development
> branch since then.
>
> Some highlights from the last year:
> - Added support for RocksDB-based content repository for better performance
> - Added SQL extension
> - Improved task scheduling
> - Various C2 improvements
> - Bug fixes and improvements to TailFile, ConsumeWindowsEventLog,
> MergeContent, CompressContent, PublishKafka, InvokeHTTP
> - Implemented RetryFlowFile and smart handling of loopback connections
> - Added a way to encrypt sensitive config properties and the flow
> configuration
> - Implemented full S3 support
> - Reduced memory footprint when working with many flow files
>
> Why 0.9.0 instead of 0.8.0?
> The codebase is starting to stabilize and we're getting closer to a 1.0
> milestone, but I feel like the extension API is not yet something we can
> commit to support in the long term. Additionally, a release branch for
> 0.8.0 has been created earlier, but it was not released, and with all the
> new development it wouldn't make sense to release it.
>
> Do you think 0.9.0 is a good target, or would you go with 0.8.0 or 1.0
> instead? Are there any blockers that should definitely make it into the
> next release?
>
> Thanks,
> Marton
>