[rabbitmq-discuss] problem about TTL time/transaction/HighAvailablity Cluster from new beginner of rabbitmq

Tim Watson watson.timothy at gmail.com
Sun May 26 14:52:01 BST 2013


On 26 May 2013, at 07:03, bupt bupt <buptloveandroid at gmail.com> wrote:
> 1.what is rabbitmq-server's default TTL time? I didn't configure any data to rabbitmq.

Infinity.

> 2.what kind of data structure and which type of data can be delivered to rabbitmq-server? 

Rabbit, like many message brokers, treats a message's payload as a binary blob of data. You can put whatever you like in there, then add (de)serialisation at the application level.

> 3.If I didn't want to lose any message , how can I design our message queue architecture.

See http://www.rabbitmq.com/reliability.html

> I know that rabbitmq has durable parameter solution but it still can't guarantee it didn't lose any message at all,

Durability has nothing to do with message persistence - its delivery mode that specifies the latter, on a per-message basis.

> especially for the data stored in memory but not in the disk. I also know that durable solution and transaction method will effect qps performance greatly, Can you help to design a good message queue system using rabbitmq and don't lose any message?

See the reliability guide I mentioned above.

> 4. Can you give me a demo code about how to use transaction of rabbitmq in java or python?
>       thank you very much ~~

http://www.rabbitmq.com/tutorials.html

Tim


More information about the rabbitmq-discuss mailing list