How does Load Balancing Work? [closed]

Question

I can't figure out how a load balancer works. Say for example on Rackspace cloud servers.

1. You have one instance with all your stuff
2. Do you clone that instance, so there are really two identical copies of it?
3. Then you get a static IP from the load balancer, and that goes to one of the two instances (which are really the same)

Is that right? Where do I figure out more of this?

Answer 1

It depends on your setup. However, the basic load balancing setup includes a single database server. In the Django Book, in the chapter about Deploying Django, there is a simple diagram that illustrates a basic load balancing configuration.

Basically, the computers that generate web pages all share the same database, which is stored on a dedicated server. Media files (e.g. CSS and images) may be stored on the same server or on a separate, dedicated media server, although this role may be performed by the load balancer itself.

However, in advanced setups aiming really high traffic or a lot of static content, you might want to serve static files on several different media servers. Finally, you may want to have a distributed database setup, but this introduces a whole new class of problems, and that is really a different topic (e.g. you can have a distributed database without using a load balancer).

Answer 2

Load balancing distributes the load onto the two instances you have. Therefore one instance isn't handling all of the work.

1. You setup a load balancer with Rackspace.
2. You receive a static IP from the load balancer.
3. You point your website's DNS to the load balancer's static IP.
4. The load balancer redirects traffic equally to the two instances.

Answer 3

There are multiple implementations of load balancing, each with their own approach that yields specific advantages based on their specializations.

1. There is typically some shared or common data inter-relating the nodes (data, user information, etc.)
2. That depends on the architecture of the application, whether it is stateful (storing data) or stateless (passing data through), there is typically common code-based shared across the nodes (which is basically "cloned" across the nodes)
3. IP assignments is typically not managed through a load balancer, but the IP configurations is usually an intimate part of its configuration (as each networked component/device will require its own IP address)

In order to understand the specifics of an implementation, the implementation itself needs to be identified.