Thanks Joel, it is clear now.
/Ivan
- Original message -
From: Joel Koshy <jjkosh...@gmail.com>
To: "users@kafka.apache.org" <users@kafka.apache.org>
Subject: Re: Kafka response ordering guarantees
Date: Thu, 18 Feb 2016 11:24:48 -0800
>
> Does this mean t
v.org> wrote:
>
>> Thanks Ben.
>>
>> As I mentioned, I'm developing a kafka library and not using standard
>> java producer.
>>
>> My question is really about protocol guarantees.
>>
>> /Ivan
>>
>> - Original message -
>&g
ol guarantees.
>
> /Ivan
>
> - Original message -
> From: Ben Stopford <b...@confluent.io>
> To: users@kafka.apache.org
> Subject: Re: Kafka response ordering guarantees
> Date: Wed, 17 Feb 2016 14:48:59 -0800
>
> So long as you set max.inflight.requests.per.c
Thanks Ben.
As I mentioned, I'm developing a kafka library and not using standard java
producer.
My question is really about protocol guarantees.
/Ivan
- Original message -
From: Ben Stopford <b...@confluent.io>
To: users@kafka.apache.org
Subject: Re: Kafka response or
So long as you set max.inflight.requests.per.connection = 1 Kafka should
provide strong ordering within a partition (so use the same key for messages
that should retain their order). There is a bug currently raised agaisnt this
feature though where there is an edge case that can cause ordering
Hello all.
I'm developing a kafka client and have a question about kafka server guarantees.
A statement from
https://cwiki.apache.org/confluence/display/KAFKA/A+Guide+To+The+Kafka+Protocol#AGuideToTheKafkaProtocol-Network
makes me a bit confused:
"The server guarantees that on a single TCP
