> > > > > > >
> > > > > > > > Thanks,
> > > > > > > > Chia-Ping
> > > > > > > >
> > > > > > > > On 2024/11/06 13:58:08 Apoorv Mittal wrote:
> > > > >
> > a
> > > > > > > voting
> > > > > > > > thread tomorrow.
> > > > > > > >
> > > > > > > > Regards,
> > > > > > > > Apoorv Mittal
> > > > > >
t; > > >
> > > > > > >
> > > > > > > On Tue, Nov 5, 2024 at 12:21 PM Apoorv Mittal <
> > > > > apoorvmitta...@gmail.com>
> > > > > > > wrote:
> > > > > > >
> > > > > > > &
gt; > > > > > >
> > > > > > > AS9: Yes, you are correct it's topic name and not topic id.
> > > > > > >
> > > > > > > AS10: Thanks, I have updated the description.
> > > > > > >
> > > >
ttal
> > > > > >
> > > > > >
> > > > > > On Tue, Nov 5, 2024 at 10:56 AM Andrew Schofield <
> > > > > > andrew_schofield_j...@outlook.com> wrote:
> > > > > >
> > > > > >> Hi Apoorv,
<
> > > > > andrew_schofield_j...@outlook.com> wrote:
> > > > >
> > > > >> Hi Apoorv,
> > > > >> A couple of additional comments.
> > > > >>
> > > > >> AS9: The tag share-partition : {group-topic-partition} confirm t
> > metric
> > > >> >> TotalFetchRequestsPerSec. That one is in the BrokerTopicMetrics
> > group
> > > >> >> so I would expect the new metric to be in the same group. This
> > comment
> > > >> >> applies
topic ID. Topic ID is
> > >> clearly more
> > >> authoritative, but I don't think it's usable in metrics.
> > >>
> > >> AS10: The description of the AcquisitionLockTimeoutMs would be better
> as
> > >> "Tracks the number of acquisitio
;>
> >> AS10: The description of the AcquisitionLockTimeoutMs would be better as
> >> "Tracks the number of acquisition locks for records which are not
> >> acknowledged within the timeout."
> >>
> >> Thanks,
> >> Andrew
> >>
> >>
gt; >> result of share session expiration or closure?
> >> >>
> >> >> AS4: I see there is FailedShareAcknowledgementRequestsPerSec.
> >> >> If this corresponds specifically to the ShareAcknowledge API, then
> >> >> the name should
10: The description of the AcquisitionLockTimeoutMs would be better as
>> "Tracks the number of acquisition locks for records which are not
>> acknowledged within the timeout."
>>
>> Thanks,
>> Andrew
>>
>> ________________
&g
m: Apoorv Mittal
> Sent: 05 November 2024 08:59
> To: dev@kafka.apache.org
> Subject: Re: [DISCUSS] KIP-1103: Additional metrics for cooperative
> consumption
>
> Hi All,
> Please let me know if there is any other feedback or suggestions.
>
> Regards,
> Apoorv Mi
a.apache.org
Subject: Re: [DISCUSS] KIP-1103: Additional metrics for cooperative consumption
Hi All,
Please let me know if there is any other feedback or suggestions.
Regards,
Apoorv Mittal
On Fri, Nov 1, 2024 at 5:30 PM Apoorv Mittal
wrote:
> Thanks Jun and Andrew for reviewing and feedbac
it would be nice to see how
>> each of them would be evident from the metrics and configurations.
>> This is going to be valuable information for operating this in production.
>>
>> AS8: As Jun mentioned, maybe a histogram would be appropriate
>> for RequestTopicPartiti
for RequestTopicPartitionsFetchRatio. Then I wonder whether it's
> worth having a separate RequestTopicPartitionsFetchEmptyCount since
> those are requests for which the fetch ratio was 0.
>
> Thanks,
> Andrew
> ________________
> From: Jun Rao
> Sen
_
From: Jun Rao
Sent: 30 October 2024 23:07
To: dev@kafka.apache.org
Subject: Re: [DISCUSS] KIP-1103: Additional metrics for cooperative consumption
Hi, Apoorv,
Thanks for the KIP. A few comments.
1. TotalShareReleaseRequestsPerSec: This is a bit weird since there is
no ShareRel
Hi, Apoorv,
Thanks for the KIP. A few comments.
1. TotalShareReleaseRequestsPerSec: This is a bit weird since there is
no ShareReleaseRequest.
2. RequestTopicPartitionsFetchRatio and InFlightBatchMessageCount: It seems
Histogram is more appropriate for them.
3. RequestTopicPartitionsFetchEmptyC
Hi Everyone,
I would like to start a discussion on KIP-1103:
https://cwiki.apache.org/confluence/display/KAFKA/KIP-1103%3A+Additional+metrics+for+cooperative+consumption
This KIP extends KIP-932 to provide additional metrics for
Queues/Cooperative consumption.
Regards,
Apoorv Mittal
18 matches
Mail list logo
