> > > >
> > > > > > > + dev@ for visibility
> > > > > > >
> > > > > > > I will investigate further today.
> > > > > > >
> > > > > > >
> > > > > > > On Wed, Jul
fect.
>> > > > > > >> >
>> > > > > > >> > I think we need to find out whether it has other symptoms
>> > > changed
>> > > > > > >> besides the performance regression to further fig
ctually I was also suspicious of the point of #snapshotState
>>> in
>>> > > > previous
>>> > > > > > discussions since it indeed cost much time to block normal
>>> operator
>>> > > > > > processing.
>>> >
;>> wangzhijiang...@aliyun.com
>>>> > > > > .invalid>
>>>> > > > > wrote:
>>>> > > > >
>>>> > > > > > Hi Thomas,
>>>> > > > > >
>>>> > > > > &g
;> > > > > > > I will investigate further today.
>>>>> > > > > > >
>>>>> > > > > > >
>>>>> > > > > > > On Wed, Jul 8, 2020 at 4:42 AM Aljoscha Krettek <
>>>>> >
---
>>>>>> > > > > > From:Thomas Weise
>>>>>> > > > > > Send Time:2020年7月17日(星期五) 05:29
>>>>>> > > > > > To:dev
>>>>>> > > > >
o verify?
> >>>>>> > > > > > [1] https://issues.apache.org/jira/browse/FLINK-17496
> >>>>>> > > > > > [2] https://issues.apache.org/jira/browse/FLINK-14881
> >>>>>> > > > > >
> >>>&g
Best,
> > Zhijiang
> >
> >
> > ------
> > From:Thomas Weise
> > Send Time:2020年7月7日(星期二) 23:01
> > To:Stephan Ewen
> > Cc:Aljoscha Krettek ; Arvid Heise <
> ar...@ververica.com>; Zhijiang
&
ssing until back pressure the source.
>> > Maybe we could dig into the process of sync execution in checkpoint.
>> E.g. break down the steps inside respective operator#snapshotState to
>> statistic which operation cost most of the time, then
>> > we might probably find the roo
ebugging if possible.
>> >
>> >
>> > Best,
>> > Zhijiang
>> >
>> >
>> > --
>> > From:Thomas Weise
>> > Send Time:2020年7月7日(星期二) 23:01
>> > To:St
From:Thomas Weise
> Send Time:2020年7月17日(星期五) 05:29
> To:dev
> Cc:Zhijiang ; Stephan Ewen ;
> Arvid Heise ; Aljoscha Krettek
> Subject:Re: Kinesis Performance Issue (was [VOTE] Release 1.11.0, release
> candidate #4)
>
> Sorry for the delay.
>
> I confirmed that the r
; You mentioned that you already reverted the SDK upgrade to verify no
> > changes. Did you also revert the [1] to verify?
> > [1] https://issues.apache.org/jira/browse/FLINK-17496
> > [2] https://issues.apache.org/jira/browse/FLINK-14881
> >
> > Best,
> > Zhijiang
> > --
>
ration of
> > sync
> > > >> checkpoint process in Kinesis sink operator, which would block the
> > > normal
> > > >> data processing until back pressure the source."
> > > >> >
> > > >> > That's a consta
upgrading amazon-kinesis-producer to 0.14.0 [1] and another is
> > for
> > > > upgrading aws-sdk-version to 1.11.754 [2].
> > > > You mentioned that you already reverted the SDK upgrade to verify no
> > > > changes. Did you also revert the [1] to verify?
&
t; > > operations
> > > > > inside KinesisProducer implementation provided by amazonaws, which
> I
> > am
> > > > not
> > > > > quite familiar with.
> > > > > But I noticed there were two upgrades related to it in
>
15 matches
Mail list logo
