Thanks for the reply.<div><br></div><div>We are just making a POC of how to use new Amazon's autoescale feature, so we can add instances based on rules. Those instances must contain rabbitmq servers in a cluster, all in the same cluster.</div>
<div>From what I read, and what you're confirming, it looks like there's no starightforward way of making it automatically.</div><div><br></div><div>Any ideas?</div><div><br></div><div>With regards <br><br><div class="gmail_quote">
On Wed, Sep 8, 2010 at 11:45 AM, Marek Majkowski <span dir="ltr"><<a href="mailto:majek04@gmail.com">majek04@gmail.com</a>></span> wrote:<br><blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex;">
Jaime,<br>
<div class="im"><br>
On Fri, Sep 3, 2010 at 11:22, Jaime Martinez <<a href="mailto:jaimemm@gmail.com">jaimemm@gmail.com</a>> wrote:<br>
> Here is the scenario:<br>
> - More than 1 pair of Active + Passive RabbitMQ servers per EC2 instance,<br>
> all of them in cluster.<br>
> - Autoescale will start a new EC2 instances and has to add the new RabbitMQ<br>
> HA servers to the same cluster as the previous instances<br>
> It looks pretty tricky since to add server instances to a cluster is<br>
> necessary to know their identifiers. Did anyone tried this? What would it be<br>
> the right way of doing it?<br>
<br>
</div>Yes, in order to join RabbitMQ nodes into a cluster you need to know the shared<br>
cookie and the addresses of all hosts in the cluster, as described here:<br>
<a href="http://www.rabbitmq.com/clustering.html" target="_blank">http://www.rabbitmq.com/clustering.html</a><br>
<br>
In my opinion clustering is not perfect. It helps scaling only for a set of use<br>
cases. Basically, to exploit the benefits you need to really know what you're<br>
doing. Generally, I wouldn't recommend it. Also, correctly configuring<br>
clustering with HA is really tricky.<br>
<br>
Saying that, I'd recommend just upgrading EC2 host if you need more capacity.<br>
Store the persistent data on EBS and when you need more power - just<br>
reattach EBS to a host with more CPU power. Sure, that's not zero-downtime,<br>
but at least it's simple and manageable.<br>
<br>
Cheers,<br>
<font color="#888888"> Marek Majkowski<br>
</font></blockquote></div><br><br clear="all"><br><br>
</div>