Electrical computers and digital processing systems: multicomput – Computer network managing – Network resource allocating
Reexamination Certificate
2000-04-08
2004-09-28
Dharia, Rupal (Department: 2141)
Electrical computers and digital processing systems: multicomput
Computer network managing
Network resource allocating
C709S203000, C709S217000, C709S219000
Reexamination Certificate
active
06799214
ABSTRACT:
BACKGROUND OF THE INVENTION
1. Field of the Invention
The present invention relates, generally, to content delivery networks and, in preferred embodiments, to systems and methods for efficient redirection of content requests to mirror sites using http meta-tag protocols to improve content delivery services.
2. Description of the Related Art
Web performance is a key point of differentiation among content providers. Crashes and slowdowns within major web sites demonstrate the difficulties companies face in trying to deal with high web traffic. As Internet backbone technologies have developed, many innovations in the area of service management have improved bandwidth and web content retrieval response time. These improvements to infrastructure, however, cannot solve traffic problems at all points within the Internet.
For example, assume in
FIG. 1
that an end user
12
in a network
14
in Japan requests access to a page from a web site
16
in a network
18
the United States. (For the purpose of simplifying the description, end users may be described herein as requesting or receiving content. However, it should be understood that the end user's Web browser actually requests or receives the content.) The request must pass through several gateways
20
,
88
, and
90
before reaching web site
16
. Although web site
16
may have large bandwidth (the ability to rapidly communicate large quantities of data), the gateways connecting the network
14
in Japan to the network
18
in the United States may be slow, and thus, when end user
12
attempts to access the page from web site
16
, the gateways may create a bottleneck. Such gateway bottlenecks may result in the access time for one page of data being on the order of 10 seconds or more. Because of the gateway bottlenecks, and because there are many uncertainties along the Internet paths from end user
12
to/from web site
16
, content delivery networks or systems are now being developed.
Fundamentally, content delivery systems are designed for two major purposes; one is to achieve load balancing, and the other is to reduce overall response time. A content delivery system may be implemented using a high speed dedicated line to deliver content while bypassing all gateways or reducing the number of Internet gateways in the transmission path. However, such a dedicated system is expensive. Another approach to implementing content delivery systems is through the use of intelligent caching, mirroring, proxy servers, or other techniques which redirect end users to available servers that contain copies of the requested content and that are close to (quickly accessible by) the end users. With some of the traffic redirected, traffic surges will decrease and end users will benefit from faster response time. The term generally used for the architecture and functionality of such networks or systems is content delivery services (CDS).
FIG. 2
is a block diagram illustrating an example of a conventional embedded object redirection content delivery system
10
. In the example of
FIG. 2
, an end user
92
located within a local network
94
, e.g., in Japan, requests access to a web page containing object
1
(obj
1
), object
2
(obj
2
), and object
3
(obj
3
) from a remote content provider original site
96
located within a network
98
in, e.g., the United States. Obj
1
, obj
2
, and obj
3
may comprise, for example, three images.
In the example system of
FIG. 2
, mirror sites
22
and
100
are employed in the Japanese network
94
. These mirror sites
22
and
100
generally contain copies of content stored in content provider original site
96
. Access to a mirror site located within the same network as the local or requesting end user
92
often results in faster access time. Thus, in the example of
FIG. 2
, when end user
92
in Japan requests data from content provider original site
96
, it would generally be much faster if end user
92
was able to access the data from a mirror site
22
or
100
in Japan, rather than accessing the image from content provider original site
96
in the United States and transferring the image across various Internet gateways. Of course, end user
92
must somehow be informed that the desired content can be accessed from a closer (e.g. more quickly accessible) source.
To facilitate this communication, content provider original site
96
maintains a cache table
28
, or accesses a cache table
28
maintained by another system, to keep track of the objects that been copied into the various mirror sites. Cache table
28
may contain a list all of the various objects or images that have been copied into other mirror sites, a mapping of the objects and the specific mirror sites that store copies of these objects, and the location of these mirror sites. In addition, cache table
28
may also contain a mapping of partial IP addresses (e.g. the most significant segments of the IP address) to the closest mirror site for those partial IP addresses.)
The “closest” mirror site is the mirror site for which response time is the fastest. Content provider original site
96
uses cache table
28
to locate a closer location from which end user
92
can access the desired content. It should be noted that although end users
92
and mirror sites
22
and
100
may be geographically close, they may be located on completely different networks. Thus, the process of selecting and copying content into a mirror site and informing end user
92
that the desired content can be found in a particular mirror site is not necessarily a simple matter of geography. Content provider original site
96
may determine the closest mirror site using the mapping in cache table
28
.
In the system of
FIG. 2
, when end user
92
first requests access to a web page within content provider original site
96
, it is in the form of a “get index.html” request
26
. Content provider original site
96
receives this request and first determines that end user
92
is from Japan because of the IP address information contained in the request. Assume, for purposes of this example, that the requested web page contains obj
1
, obj
2
, and obj
3
. Because content provider original site
96
has access to cache table
28
and also knows that requesting end user
92
is located in Japan, a customized index.html page
48
is sent back to end user
92
containing address information
50
indicating that obj
1
is located in mirror site
1
(see reference character
22
) in Japan, and that obj
2
and obj
3
are located in mirror site
2
(see reference character
100
), also in Japan.
When end user
92
receives and parses the customized index.html page 48, it determines that there are three embedded objects, obj
1
, obj
2
, and obj
3
. End user
92
then sends one “get object” request for each object. However, instead of requesting the objects from content provider original site
96
in the United States, end user
92
instead requests obj
1
from mirror site
1
, and requests obj
2
and obj
3
from mirror site
2
. In summary, content provider original site
96
modifies the index.html page sent back to end user
92
, and redirects end user
92
to alternate locations where the same object information has been mirrored for faster access. The process is transparent to the end user.
One disadvantage with the embedded object redirection content delivery system of
FIG. 2
results from the fact that only objects are mirrored. Thus, although a requested page may contain links to other pages, modified index.htnl page 48 will not contain redirection information for those pages. If a user wants to access these linked pages, a separate page request for each page must be directed to content provider original site
96
.
Furthermore, because all pages must be fetched from content provider original site
96
, some of the processing capability within content provider original site
96
must be reserved for processing all the initial requests, and preparing and transmitting customized index.html pages
48
, which could be large. This added processing may be signif
Dharia Rupal
Foley & Lardner LLP
Nguyen Quang
LandOfFree
System and method for efficient content delivery using... does not yet have a rating. At this time, there are no reviews or comments for this patent.
If you have personal experience with System and method for efficient content delivery using..., we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and System and method for efficient content delivery using... will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-3261555