incubator-cassandra-user mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From Edward Capriolo <edlinuxg...@gmail.com>
Subject Re: SchemaDisagreementException
Date Tue, 20 Nov 2012 04:54:47 GMT
http://www.acunu.com/2/post/2011/12/cql-benchmarking.html

Last I checked, thrift still had an edge over cql due to string
serialization and de serialization. Might be even more dramatic for
later columns. Not that client speed matters much overall in
cassandra's speed, but CQL client "does more".



On Mon, Nov 19, 2012 at 9:27 PM, Michael Kjellman
<mkjellman@barracuda.com> wrote:
> While this might not be helpful (I don't have all the thread history here),
> have you checked that all your servers are properly synced with NTP?
>
> From: Everton Lima <peitin.inutil@gmail.com>
> Reply-To: "user@cassandra.apache.org" <user@cassandra.apache.org>
> Date: Monday, November 19, 2012 6:24 PM
> To: "user@cassandra.apache.org" <user@cassandra.apache.org>
> Subject: Re: SchemaDisagreementException
>
> Yes I already have tested. I use the Object CassandraServer to do the
> operations instead of open conection with CassandraClient. Both of this
> object implements Iface. I think the performace of use CassandraServer
> improve because it does not open an connection, and CassandraClient (that
> use thrift) and CQL open a connection.
>
> 2012/11/19 Tyler Hobbs <tyler@datastax.com>
>>
>> Have you actually tested to see that the Thrift API is more performant
>> than CQL for your application?  As far as I know, CQL almost always has a
>> performance advantage over the Thrift API.
>>
>>
>> On Mon, Nov 19, 2012 at 1:05 PM, Everton Lima <peitin.inutil@gmail.com>
>> wrote:
>>>
>>> For some reason I can not reply my old thread in that list. So I was
>>> creating a new one.
>>>
>>> The problem is that I do not use thrift to gain in performace. Why it is
>>> nicer with concurrent modifications? I do not know why I have falling in the
>>> problem of concurrent modification if I was creating 2 keyspaces diferent in
>>> only one process with just one thread. Someone knows why?
>>>
>>>
>>> --
>>>
>>> Everton Lima Aleixo
>>> Bacharel em Ciencia da Computação
>>> Universidade Federal de Goiás
>>
>>
>>
>>
>> --
>> Tyler Hobbs
>> DataStax
>>
>
>
>
> --
>
> Everton Lima Aleixo
> Bacharel em Ciencia da Computação
> Universidade Federal de Goiás
>
> ----------------------------------
> 'Like' us on Facebook for exclusive content and other resources on all
> Barracuda Networks solutions.
> Visit http://barracudanetworks.com/facebook
>   ­­

Mime
View raw message