Help with finding examples

Hi all,
  I'm working on setting up our test environment on EC2. I'm planning to use HAProxy and stunnel together, but I can't find a lot of good up to date tutorials on how to get things set up. I have 3 nodes, each one is currently running Jetty, I would also like to install haproxy on these nodes. Our load isn't especially high at the movement, our primary goal is availability and redundancy. I will be running these instances on EC2, I cannot use Amazon's ELB, it doesn't offer enough support for firewall rules or ip assignment. Here are my requirements.

1. Sticky session support for HTTP and HTTPS
2. SSL decryption in the load balancer
3. Round robin TCP balancing for a proprietary protocol
4. Notification of failure when the current load balancer fails. Then move the EC2 elastic ip to the new elected leader.

If my understanding is correct, I would have the following module setup.

client --> 80 --> HA Proxy --> 1 of 3 jettty nodes port 8080 client --> 443 --> stunnel (With x forwarded) --> HA Proxy --> 1 of 3 jettty nodes port 8080
client --> 1000 --> HA Proxy --> 1 of 3 backend nodes port 1000

I've read the doc here.

http://haproxy.1wt.eu/download/1.4/doc/configuration.txt

I've found this tutorial

http://www.howtoforge.com/high-availability-load-balancer-haproxy-heartbeat-debian-etch

and this tutorial

http://37signals.com/svn/posts/1073-nuts-bolts-haproxy

But I have a few questions I'm struggling to find examples and information on. Any help/guidance would be greatly appreciated.

1. For sticky HTTP and HTTPS sections where ha proxy sets the cookie value, I would use "cookie proxy insert" correct?
2. We're using chef to manage our servers. Is there any event hook I can use when the current "active" proxy fails or times out? As soon as a new node becomes the primary proxy, I need to move the reserved IP address to the new node using the EC2 apis.

Thanks in advance!
Todd Received on 2011/01/08 06:24