[rabbitmq-discuss] RabbitMQ HA - Good Practices

[Mariusz Wojcik]

Hi Rom,

The load balancer is advisable - definitely do not bake in node 
hostnames/IP in your code. Read the last section, “Connecting to Clusters 
from Clients” at the bottom of the first article 
<https://www.rabbitmq.com/clustering.html>.

No, you do not need to set all queues to be mirrored. It all depends on 
your scenarios. Just bear in mind that mirrored queues have less 
performance than non-mirrored ones.

Cheers,

Mariusz
On Monday, 30 June 2014 10:14:01 UTC+1, Rom Cabral wrote:

Hi guys,
>
> Just finished reading articles about clustering and HA from the links 
> below. I would like to clarify things before I start doing some experiment.
>
> https://www.rabbitmq.com/clustering.html
> https://www.rabbitmq.com/ha.html
>
> - I have configured 3 nodes within a cluster, do I need to have a load 
> balancer to ensure it is HA? or setting the exchange & queues to be 
> mirrored enough? what is the good practice.
> - Do I need to set all queues to be mirrored?
>
> - Any other things consideration?
>
>
> ​
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.rabbitmq.com/pipermail/rabbitmq-discuss/attachments/20140630/c846f12b/attachment.html>