Web Caching Technology White Paper

Contents

Overview·· 1

Implementation· 1

Basic concepts· 1

Operating mechanism·· 2

Technical features· 3

Rich features· 3

Advanced architecture design· 3

Typical network applications· 4

Overview

Web caching is a caching technology that stores static resources and data that users have accessed in their local browser or server. This can achieve the purpose of reducing access latency and increasing access speed.

Comware uses the Web cache proxy service to cache Web resources and data. As shown in Figure 1, the Web cache proxy service is deployed on the device that resides between the Web clients and the Web servers. The device is generally located within a LAN close to the Web clients. After you deploy the Web cache proxy service on the device, the Web clients can interact with the Web servers indirectly. Instead, the Web clients send requests to the Web cache proxy. The Web cache proxy obtains resources from the Web servers uniformly and responds to the HTTP/HTTPS requests from the Web clients. Deploying the Web cache proxy service alleviates the pressure on Web servers, reduces data transmission traffic on WAN links, eases the bandwidth stress of WAN links, and speeds up website access for all users.

Figure 1 Network diagram

For both the Web clients and Web servers, the Web cache proxy service is transparent. You do not need to configure the clients and servers, and they cannot perceive the Web cache proxy.

Implementation

Basic concepts

Web cache files

Web caching stores cached content to files in the Web cache directory. When the total size of Web cache files exceeds the maximum storage space allowed, the oldest Web cache file will be overwritten.

Web cache file aging time

The Web cache file aging time is fixed at 30 days. When the device reboots or receives a request for the cached content in a file, it restarts the aging timer. If no users request the cached content before the aging timer expires, the device deletes the file.

Operating mechanism

Figure 2 Operating mechanism

As shown in Figure 2, the Web cache proxy service is deployed on the user's gateway device. The following uses HTTP access as an example to describe the operating mechanism of Web caching:

1.     After the Web cache proxy receives an HTTP GET request that a host sends to a Web server, it identifies whether it is permitted to cache the server's content. If permitted, the device will initiate Web cache processing and execute the following steps. If not permitted, the device directly forwards the request to the Web server.

2.     The Web caching process obtains the URL of the request and looks up its cache for the requested content.

¡     If the requested content is found in the cache, the device retrieves the content and sends the content to the host. The HTTP GET operation is completed.

¡     If the requested content is not found, the device re-constructs the request and sends the request to the Web server.

3.     The Web server sends the requested content to the device.

4.     The device identifies whether the Web caching feature is configured to cache the content.

¡     If yes, the device uses the Web caching feature to cache the content, re-constructs a response, and sends the response to the host.

¡     If not, the device re-constructs a response and sends the response to the host.

5.     After receiving other HTTP GET requests, the device first checks locally to see if it has been cached before. If the resource is already cached, the device returns it directly. If the resource has not been cached, the device sends a request to the Web server, caches the resource returned by the Web server, and then returns it to the client.

Technical features

Rich features

Web caching supports the following functions:

·     Secure HTTPS access—Web caching supports HTTPS access scenarios. In such a scenario, because the Web cache proxy service uses a certificate issued by a private CA, users must add the root certificate file in the Web cache directory to the trusted root certification authorities list in the client computer or browser. Otherwise, the client will prompt that the certificate is illegal.

·     Rich statistical display functions—You can view the size and download rate of files that the proxy server downloads from Web servers and files that the clients in LAN download from the proxy server at different time granularities (1 minute, 1 hour, 1 day, or 1 month).

Advanced architecture design

To fully utilize the hardware resources of the device, Comware provides the Web cache proxy service to the clients using a multiprocess parallel processing architecture. The number of processes for the Web cache proxy service is the same as the number of CPU cores on the device. The number of HTTP/HTTPS connection requests that each process can handle at the same time is only limited by the memory size and CPU capability.

The advantages of multiprocess parallel processing for the Web cache proxy service are as follows:

·     Ensure the service robustness. If one working process encounters an error, other working processes can continue to provide services normally.

·     Boost resource utilization by fully exploiting the multicore architecture of the device. A large number of concurrent HTTP/HTTPS connection requests are equally distributed to multiple cores for parallel processing.

Figure 3 Parallel service process distribution

Typical network applications

Users in a LAN typically have the same website access requirement. For example, users on an enterprise network access the same technical website to obtain the latest information. Similarly, users on a campus area network access the same recruitment website to submit resumes, and familiarize themselves with recruiting companies and available positions.

As shown in Figure 4, to increase the website access speed and alleviate the pressure on the Web servers, deploy the Web cache proxy service on the gateway device in the LAN. This allows the proxy server to cache the Web pages that users have accessed. When users in the LAN access the Web pages again, the proxy server directly responds with the locally cached information.

Figure 4 Typical Application Diagram of Web cache