+1
Nice work.
One minor nit pick: all the protocol field names are snake_case but the
proposed addition is camelCase. Let's stick with the current convention.
-Jay
On Tue, Apr 7, 2015 at 10:47 PM, Aditya Auradkar <
aaurad...@linkedin.com.invalid> wrote:
> Please vote.
> https://cwiki.apache.or
rad...@linkedin.com.invalid> wrote:
> Thanks! I've clarified the request protocol change on the KIP.
>
> Aditya
>
>
> From: Joel Koshy [jjkosh...@gmail.com]
> Sent: Wednesday, April 08, 2015 5:49 PM
> To: dev@kafka.apache.org
> Subjec
Thanks! I've clarified the request protocol change on the KIP.
Aditya
From: Joel Koshy [jjkosh...@gmail.com]
Sent: Wednesday, April 08, 2015 5:49 PM
To: dev@kafka.apache.org
Subject: Re: [VOTE] KIP-13 Quotas
Agreed on that - it is understood fro
Agreed on that - it is understood from the text of the KIP "Clients
that send current version (0) of those requests will not receive the
quota status flag in the response" => you have to send version (1) to
get the new repsonse, but it should probably be clarified on the page.
I'm +1 as well on th
Aditya,
To evolve the response protocol for produce/fetch response, we need to have
new versions of the requests as well. Version 1 of the requests will have
the same protocol as that of version 0, and will get the new response. The
rest of the stuff looks good to me.
Thanks,
Jun
On Tue, Apr 7,
Please vote.
https://cwiki.apache.org/confluence/display/KAFKA/KIP-13+-+Quotas
Thanks,
Aditya
