Win a copy of Mastering Non-Functional Requirements this week in the Design forum!
  • Post Reply Bookmark Topic Watch Topic
  • New Topic

Reverse Proxy Load Balancing?/  RSS feed

andy armstrong
Ranch Hand
Posts: 154
  • Mark post as helpful
  • send pies
  • Quote
  • Report post to moderator
On the mock exam on
The answer to the question
What is reverse proxy load balancing?

(D) A technique used to target certain requests to certain backend servers, e.g. All Servlet requests from one server. All static HTML from another.
However after doing some reading I "think"
that Reverse proxy load balancing is used to
avoid the 'N' officially assigned IP addresses and DNS entries for the back-end servers and thus
support fault tolerance and/or sever fall over.
The reverse proxy can also split up requests
according to load and in Apache's case to bandwith..
Thus the answer for the above question would be..
Splitting requests evenly amoungst all back end servers
Any comments would be appreciated..
Arun Rajamani
Posts: 4
  • Mark post as helpful
  • send pies
  • Quote
  • Report post to moderator
I have a feeling that Ian's answer is correct in the mock test .........
The "Reverse-proxy" model in load balancing uses the caching features of a proxy server. The usual configuration is to have a reverse-proxy server pool before the web/app server that has to be accessed.
The proxy server pool caches HTML docs. Only non-cached (servlet or CGI requests) are routed to the web/app server.
Any client trying to access cached HTML docs will be serviced by the proxy server pool.
All other requests will be sent thro' the proxy server and serviced by the web/app server.
So in a way, the proxy server targets certain requests across servers.
A DNS-round robin server may also be used in front of the reverse-proxy pool.

It is sorta covered in the JavaRanch Style Guide.
  • Post Reply Bookmark Topic Watch Topic
  • New Topic
Boost this thread!