Electrical computers and digital processing systems: multicomput – Remote data accessing
Reexamination Certificate
1996-03-28
2003-07-29
Dinh, Dung C. (Department: 2153)
Electrical computers and digital processing systems: multicomput
Remote data accessing
C709S227000, C709S229000
Reexamination Certificate
active
06601091
ABSTRACT:
BACKGROUND OF THE INVENTION
The Internet is a large scale wide-area network connecting a rapidly growing number of sites. The Internet consists of a communications protocol and addressing scheme allowing any two computers on the Internet to communicate with each other. This backbone is implemented by several layers allowing specific types of communication between a wide variety of systems. File Transfer Protocol (ftp) capable Internet sites respond to a particular type of communications request by exposing a list of files and directories. Hyper Text Transfer Protocol (http) capable Internet sites provide access to a specific document which contains text formatted with predefined formatting commands which lay out the format of the text and includes pointers to other http documents or graphic images.
Several other protocols exists, and others may develop as the Internet grows. The present invention is directed to ftp and http sites, but is intended to handle other protocols as well. The term “World Wide Web”, or simply “Web”, refers collectively to the collection of Internet sites responding to these protocols, and specifically to the collection of http-compatible sites.
In order to access the Internet, a user must have access to a direct or indirect connection to the Internet hardware backbone. The backbone is provided by a number of large Internet sites at private service provider companies, universities and government institutions. These sites accept data destined for some specific Internet address and route the data packet to the destination in-accordance with a predefined routing protocol.
Smaller Internet sites can be part of the routing mechanism, thus providing a hierarchical network of continuously decreasing bandwidth. At the bottom end, individual end-users access a local site through a modem or other connection. The individual end-user's computer is then considered to have its own Internet address and data can be “routed” to the end-user's computer. Small networks can also be connected to the Internet through several methods, allowing all users on a small network to access the Internet.
Computers connected to the Internet can be “Servers” which generally respond to ftp or http requests, and/or “Clients” or “Browsers” which primarily let users access information provided by a Server. Such Browser software requests a site address from the user, and accesses the site, presenting the user with whatever information is made available by the Server at the site the user selected.
FIG. 1
shows the general operation of an Internet Browser. In
101
the user requests access to a particular site (either directly or by defining a “home” site that is always displayed first). In
102
the Browser looks up the Internet protocol (ip) address of the requested data and contacts the remote sites. In
103
the data is requested. It is retrieved at
104
and presented to the user at
105
.
For ease of use, users are not required to enter the actual numeric Internet address (or ip address) of the site which they are interested in contacting. Instead, the Internet contains, at various service provider locations, Domain Name Servers (DNSs). These DNSs contain databases of Internet addresses and names of the sites that provide the resource associated with the name. These names stored in the DNS databases conform to a specification called the Uniform Resource Location (URL) specification. Thus, the user need only know the URL name of the site they wish to access, and the Browser software will search a local DNS for the actual address of the site so named.
A known method (see Mosaic Web Browser or Netscape version
1
.
0
) includes the ability for caching documents that have been accessed previously, whether in the same or a previous session. (The CompuServe Interface Manager for Windows, WinCIM, also provides a cache of certain bitmaps and documents). This method allows subsequent accesses to the same document, and may, prior to retrieving a document, check the local cache to see if the document exists locally and if it has been modified since it was copied locally.
FIG. 2
diagrams the operation of this method. In
201
the user indicates which site to contact, as usual. In
202
the site is contacted; the Browser determines whether the data has been modified via a last modified time stamp, checksum or other procedure. In
203
the Browser determines whether the modified version of the data is available locally. If not, the data is requested (
204
), retrieved (
205
) and copied into the local cache (
206
). The appropriate data, either the local or retrieved copy, is then displayed (
207
).
This is a useful methodology and can be implemented together with the present invention. This technique, however, only provides faster access to documents that have been accessed previously and, have not subsequently changed. Thus, this method is unable to provide an improvement in accessing data that changes frequently. Also, this technique, if it is applied across multiple sessions, tends to consume large quantities of local storage for the document cache. In order to provide any improvement, the local storage for cross-session access must be sufficient to provide a meaningful portion of what the user may see, and thus the method is not generally practical.
In another prior art method (see Dr. Dobb's Journal April 1996, The Harvest Object Cache and references therein) the problem of speeding up the overall response time of the Internet and reducing the load on the Internet is tackled. This is accomplished via distributed caching at the local or regional network level.
FIG. 5
is a simplified diagram of this system. Local networks
503
are connected to a regional network
502
which in turn is connected to a wide area network backbone
501
. A typical data request (
506
) is initiated at a local workstation, and flows through a local regional network to the network backbone and then to another regional network and finally to another local network. The response would follow the reverse path back to the user.
In this method, cache storage
504
is added at various points in the network, and frequently-accessed documents are kept in duplicate in the cache storage and accessed through data path
505
. The response to a data request is thus quicker when the requested data can be retrieved from the cache
504
, as the request need not be passed down to the appropriate local network and back up again.
This is a useful and laudable goal. However, the end-user is interested in the apparent performance of the end-user's own workstation, and in the time spent actually connected to the Internet. So, the actual response time of the Internet is less material if it seems to the user that the Internet response was instantaneous. In other words, the time it took the user's workstation to access the information is immaterial provided it appears as soon as it is requested by the user. Furthermore, since the user is charged for connect time rather than message units, the user is concerned with the overall time spent logged on to the Internet rather than the amount of time spent accessing particular objects.
This prior art caching methodology does not satisfy three needs: First, it is of little assistance in the accessing of documents that change frequently (for instance, newspaper front pages) or to documents that are accessed infrequently or have never been accessed previously, as such documents would not be in a cache. Second, this method can be difficult to deploy, requiring some cooperation between the regional networks, and increases the cost of providing network services. In addition, this method does not improve access speed if a low bandwidth bottleneck exists between the user's workstation and the cache which contains the requested data.
Because of the varying bandwidths of the various sites on the Internet and because of the indefinite nature of the connection between servers and browsers, there can be a varying degree of lag time between an access request and a response. The comb
Dinh Dung C.
Hamaty Christopher J.
Networks Associates Technology Inc.
Silicon Valley IP Group, LLC
Zilka Kevin J.
LandOfFree
Method for improving the perceived performance of a computer... does not yet have a rating. At this time, there are no reviews or comments for this patent.
If you have personal experience with Method for improving the perceived performance of a computer..., we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Method for improving the perceived performance of a computer... will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-3099128