Patent Application
20050165931
1. Field of the Invention
This invention relates to communication. Specifically, the present invention relates to content access.
2. Description of the Prior Art
Internet technology is pervasive and widely deployed. A large variety of content may be accessed over the Internet. The content is often stored in servers. A client machine communicates with a server to access the content.
As the Internet continues to expand, a large volume of clients attempt to access content on the Internet. This may result in a bottleneck if the clients are each attempting to access content from the same server. As a result, in conventional systems, the same content is often deployed on multiple servers. In addition to enabling more clients to get simultaneous access to the data, deploying the content on multiple servers often has ancillary benefits, such as system redundancy, greater security, etc.
However, the server is not the only bottleneck. The communication path from the client to the server may also serve as a bottleneck for communication. In addition, while there are some techniques for load balancing, the primary server is often disproportionately loaded relative to the other servers. As a result, the client experiences a slow response time either because of the disproportionate loading of a server or because of bottlenecks in the communication path.
Thus, there is a need for a method of discerning which server will provide the best response time to a client. There is the need for a method of determining which communication path will provide the best response time and throughput to a client. Lastly, there is a need for a method of determining which client in combination with the communication path will provide the best response time to a client.
A method is implemented that determines which server in a sequence of servers will provide the best response time for a client accessing the server. A bandwidth probe is implemented. In one embodiment, the bandwidth probe provides a mechanism for testing the server response and the communication path throughput to determine which server and/or communication path will have the best response time. As a result, the probe provides a quick and accurate way of measuring the response time of the server in combination with the communication path.
A method of communicating comprises the steps of receiving a communication from a client; instructing at least one server to begin a bandwidth probe in response to receiving the communication from the client; receiving results of the bandwidth probe in response to instructing the at least one server; and sending a redirect message to the client in response to receiving the results of the bandwidth probe.
A method of communicating comprises the steps of receiving a start packet; receiving a train of consecutive packets; receiving an end packet; computing time dispersion in response to receiving the start packet, receiving the train of consecutive packets, and receiving the end packet; and communicating a result in response to computing the time dispersion, wherein a server is selected for access in response to communicating the result.
A method of accessing a server comprises the steps of receiving an access request from a client; instructing a plurality of servers to each operate a bandwidth method in response to receiving the access request, the bandwidth method determining available bandwidth; receiving a bandwidth indication from each of the plurality of servers; selecting an identified server in response to receiving the bandwidth indication from each of the plurality of servers; and redirecting the client to the identified server.
While the present invention is described herein with reference to illustrative embodiments for particular applications, it should be understood that the invention is not limited thereto. Those having ordinary skill in the art and access to the teachings provided herein will recognize additional modifications, applications, and embodiments within the scope thereof and additional fields in which the present invention would be of significant utility.
Servers 100 and 102 are in communication with network 104. Information, such as content, is communicated across network 104. In one embodiment, network 104 is implemented as a packet-switching network. In a second embodiment of the present invention, network 104 may be implemented as a circuit-switching network. In yet another embodiment of the present invention, network 104 may be implemented as an integrated packet and circuit switching network, a Local Area Network, a Wide Area Network, a wireless network, etc.
A media controller 106 is in communication with network 104. The media controller 106 represents any entity capable of controlling access to servers 100 and 102. The media controller 106 may be implemented in software, hardware, or in a combination of software and hardware.
A client 108 is in communication with the network 104. The client 108 represents a device used by an end-user to access resources in the network 104. The client 108 may be implemented in software, hardware, or in a combination of software and hardware.
Input devices, such as tactile input device, joystick, keyboards, microphone, communication connections, or a mouse, are shown as 212. The input devices 212 interface with the system through an input interface 214. Output devices, such as a monitor, speakers, communications connections, etc., are shown as 216. The output devices 216 communicate with computer architecture 200 through an output interface 218.
At 302, the media controller 106 communicates with each server (100, 102) that has the content to begin a bandwidth probe. At step 304, the servers contacted by the media controller 106, such as servers 100 and 102, each launch software and/or hardware that probe communication bandwidth (i.e., bandwidth probe). Once the servers 100 and 102 have completed the bandwidth probe, each server 100 and 102 communicate the results of the bandwidth probe to the media controller 106 as stated at step 306. At step 308, the media controller 106 selects a server (100, 102). The media controller 106 may select the server 100 or 102 based on a variety of criteria. At step 310, the media controller 106 sends a redirect message to the client 108. The redirect message identifies which server 100 or 102 that the client 108 should use. As a result of the redirect message communicated at step 310, the client 108 contacts the identified server (100, 102) as stated at 312. At 314, the identified server 100 or 102 communicates the content to client 106.
In one embodiment of the present invention, a bandwidth probe consists of a short “train” of packets transmitted at the speed of the outgoing interface to a given endpoint. The endpoint reports back on the arrival time of the start and end of the train. The spacing between the packets at the receiving endpoint is reflective of the available bandwidth along the path, allowing an estimate to be formed. There are several different variants of the bandwidth probe depending on the environment of the endpoint being probed. The variations of the bandwidth probe differ in the method used to record the time differences between packets in the train. For example, four variations of the bandwidth probe are presented: (1) an Internet Control Message Protocol (ICMP) echo with ICMP Timestamp Record may be used, (2) ICMP echo with Internet Protocol (IP) Timestamp Record may be used, (3) Transmission Control Protocol (TCP) Push/Reset with sender-based time recording may be used, and/or (4) ICMP echo with sender-based time recording may be used. It should be appreciated that although four bandwidth probe variations have been defined and described, other variations of bandwidth probes may be implemented and are within the scope of the present invention.
Referring to
At step 404, end-of-train packets are sent. For example, end-of-train packets are sent from servers 100 and 102 to client 108. In one embodiment, the end-of-train packets are defined by one of the bandwidth probe variations. At 406, the servers 100 and 102 receive roundtrip packets communicated from the client 108. In one embodiment, the roundtrip packets are time stamped, for example, the receivers' timestamp the roundtrip packets. At 408, the servers 100 and 102 then use the roundtrip packets to compute bandwidth performance measures, such as throughput, delay, and packet loss.
In
In one embodiment of the present invention, the train-of-packets 500, 502, 504, 506, 508, 510, 512, 514, 516, and 518 are used to represent start-of-train packets, a train of N consecutive packets, and end-of-train packets. The start-of-train packets and the end-of-train packets are time stamped at the client machine 507 and then again at the server machine 501. Using the start-of-train packets and the end-of-train packets at the client machine 507, the receiver time dispersion shown as 520 may be calculated. In addition, using the start-of-train packets and the end-of-train packets at the server machine 501, the sender time dispersion 522 may be calculated. For example, in the case where a time stamp is implemented, the time stamp may be used to calculate dispersion. Further, using the receiver time dispersion 520, number-of-packets sent and size-of-the-packets throughput may be calculated. In the case where the timestamp function is not available on the receiving node, the sender time dispersion 522 is used to calculate the throughput. Round trip delay and packet loss may also be calculated.
While the present invention is described herein with reference to illustrative embodiments for particular applications, it should be understood that the invention is not limited thereto. Those having ordinary skill in the art and access to the teachings provided herein will recognize additional modifications, applications, and embodiments within the scope thereof and additional fields in which the present invention would be of significant utility.
It is, therefore, intended by the appended claims to cover any and all such applications, modifications, and embodiments within the scope of the present invention.
