[rabbitmq-discuss] what's the difference between virtual host and clustering

Jerry Kuch jerryk at vmware.com
Wed Aug 3 02:33:09 BST 2011

Hi, Yu-Teh: 

The similarity you're speaking of doesn't run especially deep, and might even mislead you if you internalize it too thoroughly. 

You can, as you say, have many virtual hosts within a RabbitMQ server, but this is independent of the fact that you can cluster multiple RabbitMQ servers into a cooperating whole. 

Clustering involves the management of certain shared state across all of the Rabbit servers that participate in a cluster, but not all server state is shared (for example a given queue will exist only on one of the Rabbit servers in a cluster in the current implementation while the metadata used for maintaining users, exchanges, queues, bindings, and the like however is shared through Erlang's Mnesia database). 

Best regards, 

----- Original Message -----

From: "Yu-Teh Shen" <shenyute at gmail.com> 
To: rabbitmq-discuss at lists.rabbitmq.com 
Sent: Tuesday, August 2, 2011 5:52:16 PM 
Subject: [rabbitmq-discuss] what's the difference between virtual host and clustering 

According to the AMQP 0.9.1, the virtual host is a data partition within the server and it comprises its own name space, a set of exchanges, queues ..etc. 

I thought it just very similar to the clustering which combine many erlang node together. 

Or can I say that they are under different layer of abstraction. Once you combine many rabbitmq erlang node together, it is called cluster. 

And if you want to isolate no matter a single node or cluster into different partition, it is called virtual host. 

Is my understanding right? 


Shen, Yu-Teh 

rabbitmq-discuss mailing list 
rabbitmq-discuss at lists.rabbitmq.com 

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.rabbitmq.com/pipermail/rabbitmq-discuss/attachments/20110802/012d9a16/attachment.htm>

More information about the rabbitmq-discuss mailing list