You are viewing a plain text version of this content. The canonical link for it is here.
Posted to users@kafka.apache.org by Hamza HACHANI <ha...@supcom.tn> on 2016/10/13 13:38:42 UTC

Problem with ke-value in the KvStorre

Hi,

I've designed two processors with a different topology.

The issue is that in the firs  topology in one node i was able to associate diffrent messages (key,value) where the key could be the same and so i was able to do something like countbyKey.

In the second example when i liked to do the same i noticed that this was not possible. eny new value assiated to a key is erased by the next value so there is a respect to the unicity of the key.

I think this is really weird.

Does anybody has an explenation or a suggestion ?

Thanks in advance,


Hamza

RE: Problem with ke-value in the KvStorre

Posted by Hamza HACHANI <ha...@supcom.tn>.
Sorry i was saying anything.

Please consider  as if i didn't say anything.

Kafka does ensure the unicity of the key.

________________________________
De : Hamza HACHANI <ha...@supcom.tn>
Envoyé : jeudi 13 octobre 2016 01:38:42
À : users@kafka.apache.org
Objet : Problem with ke-value in the KvStorre

Hi,

I've designed two processors with a different topology.

The issue is that in the firs  topology in one node i was able to associate diffrent messages (key,value) where the key could be the same and so i was able to do something like countbyKey.

In the second example when i liked to do the same i noticed that this was not possible. eny new value assiated to a key is erased by the next value so there is a respect to the unicity of the key.

I think this is really weird.

Does anybody has an explenation or a suggestion ?

Thanks in advance,


Hamza