Hi, all. It seems the format is not normal. So I add a google doc in
link[1]. This discussion is about the interfaces in FLIP-95: New Table
Source And Table Sink and propose to merge two interfaces
SupportsWatermarkPushDown and SupportsComputedColumnPushDown.
I am looking forward to any opinions
Seth is right, I was just about to write that as well. There is a
problem, though, because some of our TypeSerializers are not
deterministic even though we use them as if they were. Beam excludes the
FloatCoder, for example, and the AvroCoder in certain cases. I'm pretty
sure there is also weir
Hi all,
To complete the Source refactoring work, I'd like to revive this
discussion. Since the mail thread has been dormant for more than a year,
just to recap the motivation of the FLIP:
1. The FLIP proposes to standardize the connector metrics by giving
guidance on the metric specifications, in
Hi timo,
Got it, thanks for your reply
在 2020年9月4日星期五，Timo Walther 写道：
> Hi Jun,
>
> you don't need contributor permissions to contribute to Apache Flink.
>
> Just open/pick an issue and ping a committer to assign you to this ticket.
>
> Regards,
> Timo
>
> On 04.09.20 14:28, jun su wrote
There is already an implicit assumption the TypeSerializer for keys is
stable/deterministic, RocksDB compares keys using their serialized byte
strings. I think this is a non-issue (or at least it's not changing the
status quo).
On Fri, Sep 4, 2020 at 6:39 AM Timo Walther wrote:
> +1 for getting
Hi Jun,
you don't need contributor permissions to contribute to Apache Flink.
Just open/pick an issue and ping a committer to assign you to this ticket.
Regards,
Timo
On 04.09.20 14:28, jun su wrote:
Hi Guys,
I want to contribute to Apache Flink.
Would you please give me the permission as a
Hi Guys,
I want to contribute to Apache Flink.
Would you please give me the permission as a contributor?
My JIRA ID is sujun1020.
--
Best,
Jun Su
Andrey Zagrebin created FLINK-19142:
---
Summary: Investigate slot hijacking from preceding pipelined
regions after failover
Key: FLINK-19142
URL: https://issues.apache.org/jira/browse/FLINK-19142
Proj
+1 for getting rid of the TypeComparator interface and rely on the
serialized representation for grouping.
Adding a new type to DataStream API is quite difficult at the moment due
to too many components that are required: TypeInformation (tries to deal
with logical fields for TypeComparators),
I like the proposal! I didn't check the implementation section in detail
but the SQL DDL examples look good as well as the options for specifying
how fields are mapped to keys/values look good.
Aljoscha
On 04.09.20 11:47, Dawid Wysakowicz wrote:
Hi Timo,
Thank you very much for the update. I
Hi Xuannan,
thanks for the update. Here is some final feedback from my side. Maybe
others have some final feedback as well before we continue to a voting?
Some mistakes that we should fix in the FLIP:
- The FLIP declares `Table cache();` but I guess this should be
`CachedTable cache();` now.
Yunhui created FLINK-19141:
--
Summary: Flink Job Submitted on Yarn Does not Work
Key: FLINK-19141
URL: https://issues.apache.org/jira/browse/FLINK-19141
Project: Flink
Issue Type: Bug
Compo
Thanks for publishing the FLIP!
On 2020/09/01 06:49:06, Dawid Wysakowicz wrote:
> 1. How to sort/group keys? What representation of the key should we
> use? Should we sort on the binary form or should we depend on
> Comparators being available.
Initially, I suggested to Dawid (in priva
Hi Timo,
Thank you very much for the update. It indeed covers the full story in
more details. I agree with the proposal.
On 04/09/2020 10:48, Timo Walther wrote:
> Hi everyone,
>
> I completely reworked FLIP-107. It now covers the full story how to
> read and write metadata from/to connectors and
Benchao Li created FLINK-19140:
--
Summary: Join with Table Function (UDTF) SQL example is incorrect
Key: FLINK-19140
URL: https://issues.apache.org/jira/browse/FLINK-19140
Project: Flink
Issue Ty
Hi everyone,
I completely reworked FLIP-107. It now covers the full story how to read
and write metadata from/to connectors and formats. It considers all of
the latest FLIPs, namely FLIP-95, FLIP-132 and FLIP-122. It introduces
the concept of PERSISTED computed columns and leaves out partition
Mohammad Hossein Gerami created FLINK-19139:
---
Summary: flink-conf.yml read values from environment variables
Key: FLINK-19139
URL: https://issues.apache.org/jira/browse/FLINK-19139
Project: F
@Dawid
Thanks for the information and preparing a fix for FLINK-19133!
I have made it a blocker for 1.11.2 and will keep tracking its status.
@Till
Thanks for the updates and efforts for FLINK-18959!
Thanks,
Zhu
Till Rohrmann 于2020年9月4日周五 下午3:41写道：
> Fyi, FLINK-18959 has been merged into the r
Fyi, FLINK-18959 has been merged into the release-1.10 branch.
Cheers,
Till
On Thu, Sep 3, 2020 at 2:38 PM Dawid Wysakowicz
wrote:
> User has just reported another issue FLINK-19133 which I think should be
> a blocker for the 1.11.2 release. I'll try to prepare a fix as soon as
> possible.
>
>
+1
Thanks,
Zhu
Xintong Song 于2020年9月4日周五 下午1:26写道：
> Thanks for starting this vote.
>
> +1 from my side.
>
> Thank you~
>
> Xintong Song
>
>
>
> On Fri, Sep 4, 2020 at 12:37 AM Till Rohrmann
> wrote:
>
> > Hi devs,
> >
> > I'd like to start a voting thread on FLIP-138 [1], which proposes to ma
20 matches
Mail list logo
