Configuring Domain name and IP in a Virtual server

[jayanthij]

I have been using F5 load balancer for configuring Virtual servers and feel that riverbed Stingray Traffic manager has totally a different way of maintaining virtual servers.

Havent been successful in configuring an external domain name and ip address to a virtual server i have created.

It throws error saying - IP address is not in local address space. I suspect that the issue is to add the subnet somewhere in stingray to recognize the external IP address space and not able to find where to.

Not finding any proper documentation on it as well. Can someone direct me in the right direction here? 

Appreciate any help on this, Thank you!

[owen]

Hi - Stingray does have a different way of configuring virtual servers.  It's configured in a similar fashion to most other IP endpoints for traffic, e.g. a webserver.

Stingray should be the endpoint for the incoming traffic, meaning that it should own (raise) the destination IP address.  Virtual servers are configured to bind to selected IP addresses and ports.

For example, if you wanted to manage traffic for www.example.com (131.31.1.3), you would generally need to raise 131.31.1.3 on the Stingray Traffic Manager (typically within a traffic IP group rather than as a static IP address).

Sometimes, Stingray will be deployed within a DMZ and the public IP address (131.31.1.3) will be NAT-ted to an internal one at the edge of the DMZ; then you'll find Stingray listening on the internal IP.

Basically - you will need to map the DNS name of your service to an IP address that Stingray is managing.

regards

Owen

[jayanthij]

But how to find what is the IP address space that the Stingray manages?

[owen]

Stingray will only manage traffic that is directed to it, i.e. to an IP address that the Stingray server has raised*.

A typical deployment of Stingray will have two or three network interfaces; one interface on the public side (where traffic is recieved), one on the server side (generally a private network for the webservers) and optionally an additional interface for management traffic.  You can of course use fewer or more interfaces as you require.

Generally, your DNS (www.example.com) will resolve to an IP address that sits on the same subnet as Stingray's public interface.  In the simple case, you can configure that address as a static IP address on the Stingray server.

More commonly, you would configure that IP address as a 'Traffic IP address' and the Stingray cluster will raise that address on the public interface for you, and will re-raise it on another cluster member if it is dropped.  That's how the 'single hosted' fault-tolerance mode functions.

"Where to find the IP address range that Stingray manages?" - make sure that it's deployed with an interface on the desired public subnet.

* Footnote: it's possible to get Stingray to manage traffic to other IP addresses and port ranges, using iptables on the host server to intercept and rewrite traffic to a local address.  This is less common, so don't plan to use it unless absolutely necessary.