Question

Responsibility of a server is to get request and giving appropriate response to the request. Then, what is the responsibility of a proxy server? Is there any relationship between server and proxy server(i.e, if server gets overloaded, then proxy server will pretend like a server)? If it is not true, then what is the solution when the server is getting overloaded?

Answer 1

In general Serge’s answer is correct and succinct, and I upvoted it accordingly.

To expand on Serge’s definition a bit… because your question talks about the relationship between servers, it sounds like you might be thinking about a category of proxy server called a reverse proxy.

First let’s quickly define both forward and reverse proxies. Generally speaking, when using “forward” proxies, the client knows about the target server, and asks the proxy server to forward the request to that known server. In contrast, when using a “reverse” proxy, the client does not know what the target server is. The reverse proxy is set up to invisibly forward the request to a behind-the-scenes server, which does the actual work, and the reverse proxy passes along the response.

Reverse proxies are very commonly used, for a variety of reasons:

1. Hiding the actual servers from direct Internet access improves your security.
2. You can “terminate” SSL at the proxy, which means decrypting the request at that point, so the web servers don’t have to perform that computationally expensive task.
3. The proxy can act as a “load balancer”, distributing requests among a pool of servers doing the work, so your service can scale better.
4. You can cache content at the proxy, greatly speeding up responses to duplicate requests.
5. You can serve static content very fast, but still mix in dynamically-generated content that is produced by application servers.

Hopefully this clarifies the relationship between “servers” and “proxy servers”.