[rabbitmq-discuss] RabbitMQ dying during evaluation testing

Alexis Richardson alexis.richardson at gmail.com
Wed Apr 22 14:36:43 BST 2009


Yes, give that a try and let us know how you get on!


On Wed, Apr 22, 2009 at 2:28 PM, Jacek Furmankiewicz <jacek99 at gmail.com> wrote:
> Ah, I missed that. I guess that relates to this part of the docs?
>
> channel.basicPublish(exchangeName, routingKey,
>                      MessageProperties.PERSISTENT_TEXT_PLAIN,
>
>                      messageBodyBytes);
>
> channel.basicPublish(exchangeName, routingKey,
>                      new AMQP.BasicProperties
>                        (contentType, contentEncoding, headers, deliveryMode,
>
>                         priority, correlationId, replyTo, expiration,
>                         messageId, timestamp, type, userId,
>                         appId, clusterId),
>                      messageBodyBytes);
>
> Let me try that, thanks for pointing that out.
>
> Cheers, Jacek
>
>
> On Wed, Apr 22, 2009 at 9:11 AM, Alexis Richardson
> <alexis.richardson at gmail.com> wrote:
>>
>> Jacek,
>>
>> Is your requirement:
>>
>> 1. to persist messages on disk so that you do not lose them in a crash?
>> 2. to remove messages from memory when they are stored on disk?
>>
>> RabbitMQ already does 1.  If you make your messages persistent, they
>> are stored *both* in memory *and* on disk.
>>
>> As Matthias mentioned, we are working on 2, which is needed in some
>> situations.
>>
>> alexis
>>
>>
>>
>>
>>
>>
>> On Wed, Apr 22, 2009 at 1:01 PM, Matthias Radestock <matthias at lshift.net>
>> wrote:
>> > Jacek,
>> >
>> > Jacek wrote:
>> >> Understand. Is it just a matter of switching from RAM to disk-based
>> >> tables in
>> >> Mnesia?
>> >
>> > Mnesia is only used for routing and config information. It's not
>> > designed to handle the kind of access patterns and volumes required for
>> > message storage.
>> >
>> >> If you could enhance it to
>> >> start backing up at least some of the queued up messages to disk to
>> >> ensure
>> >> reliability it would go a long way for us.
>> >
>> > Understood. We are working on that, but it's a complex piece of work and
>> > hence will take a while.
>> >
>> >
>> > Matthias.
>> >
>> > _______________________________________________
>> > rabbitmq-discuss mailing list
>> > rabbitmq-discuss at lists.rabbitmq.com
>> > http://lists.rabbitmq.com/cgi-bin/mailman/listinfo/rabbitmq-discuss
>> >
>
>




More information about the rabbitmq-discuss mailing list