preview

Designing A Surrogate Server And Its Numerical Ip Address

Decent Essays

In this approach, the DNS performs the mapping between a surrogate server’s symbolic name and its numerical IP address.
It is used for full-site content selection and delivery.
In DNS-based request-routing, a domain name has multiple IP addresses associated to it. When an end-user’s content request comes, the DNS server of the service provider returns the IP addresses of servers holding the replica of the requested object. The client’s DNS resolver chooses a server among these. To decide, the resolver may issue probes to the servers and choose based on response times to these probes. It may also collect historical information from the clients based on previous access to these servers.

Advantage:
 DNS-based approach is extremely popular …show more content…

URL rewriting
Though most CDN systems use a DNS based routing scheme, some systems use the URL rewriting. It is mainly used for partial-site content selection and delivery where embedded objects are sent as a response to client requests.

In this approach, the origin server redirects the clients to different surrogate servers by rewriting the dynamically generated pages’ URL links. For example, with a Web page containing an HTML file and some embedded objects, the Web server would modify references to embedded objects so that the client could fetch them from the best surrogate server.

To automate this process, CDNs provide special scripts that transparently parse Web page content and replace embedded URLs

URL rewriting can be pro-active or reactive.
 In the pro-active URL rewriting, the URLs for embedded objects of the main HTML page are formulated before the content is loaded in the origin server.
 In reactive approach involves rewriting the embedded URLs of a HTML page when the client request reaches the origin server.

Advantage:
The main advantage of URL rewriting is that the clients are not bound to a single surrogate server, because the rewritten URLs contain DNS names that point to a group of surrogate servers. Also finer level of granularity can be achieved through this approach since embedded objects can be considered

Get Access