1. Field of the Invention
The present invention relates generally to traffic redirection in a distributed system. More particularly, the present invention relates to a method and apparatus for redirecting traffic to a particular network location.
2. Description of the Related Art
A computer network may be defined as an interconnected collection of autonomous computers. In a distributed system, the existence of these multiple autonomous computers is transparent to the user. To achieve this transparency, allocation of jobs to processors and all other system functions must be automatic. These automated system functions are typically provided by an operating system.
In general, the operating system hides the details of the hardware from the user and provides the user with a convenient interface for using the system. More particularly, the operating system is responsible for allocating resources within the distributed system and schedules the execution of various services accordingly. Thus, the operating system selects the best processor, locates and transfers all corresponding service requests to the appropriate location. In this mariner, the operating system ensures that system resources such as file servers are used in an efficient manner.
The resource allocation provided by the operating system includes the retrieval and processing of data. Often, this data is stored on one or more shared file servers. Users in such a system are called clients. Communication from a client generally comprises a request message asking for a particular service to be performed. The service request message is then sent to an appropriate server. The server then does the work requested and sends back a reply. Thus, data is accessed and processed by the server in accordance with the service request message.
Similarly to allocating resources among multiple servers in a simple network, it is also important to appropriately allocate client requests among servers on the Internet. For instance, redundant data centers are often required in order to provide a backup in the event of failure of one of the data centers. Thus, when there are globally distributed data centers that provide the same content, various metrics are often used to select the optimum site to which to send a client request.
When a client requests access to a particular Internet domain name on the Internet, this domain name must be translated to an Internet Protocol (IP) address. More particularly, the Domain Name System (DNS) is used to map domain names of hosts and clients on the Internet to their corresponding IP addresses. Since a central list of domain name-IP address mappings would be impractical, such lists are distributed throughout the Internet in various DNS servers. By way of example, a client typically sends a DNS query to a DNS server that includes a host name with an indication that an IP address is requested. The DNS server then returns an IP address associated with the host name.
Ping is a basic Internet program that enables a client to verify that a particular IP address exists and that the host can accept requests. More particularly, ping is typically used diagnostically to ensure that a host computer is actually operating. Ping is also used with a host that is operating to determine how long it takes to get a response back. Ping operates by sending a packet to a designated address and waiting for a response.
Although the round-trip-response time metric is a valuable tool that may be used to determine the most efficient route and therefore the most appropriate web server to send a request, there are several drawbacks to this approach. First, the ping packet is often blocked by firewalls. Second, another method using TCP syn ack packets may look like a security denial of a service attack. As a result, another “boomerang” race system is also commonly used to determine the optimum route for routing of client requests.
Both the round-trip-response time metric and the boomerang race system methods are valuable in determining a fastest link among multiple links. However, such calculations rely upon the premise that these travel times are static. In other words, these calculations do not take into consideration that transmission times rely, at least in part, upon network availability of service providers of the Internet. Accordingly, neither of these methods is capable of determining a fastest provider from among multiple facilities or service providers. It would therefore be desirable if a mechanism were provided to enable the optimum service provider to be selected dynamically among multiple service providers on a per transaction basis.
While the Domain Name System is important in locating a particular web site or domain name on the Internet, the Hypertext Transfer Protocol (HTTP) is important in exchanging files on the World Wide Web. Each web server contains an HTTP daemon that is designed to wait for HTTP requests and handle them when they arrive. A web browser is an example of a HTTP client that sends requests to server machines. When the browser user enters a file request by “opening” a web file (e.g., by typing in a Uniform Resource Locator) or clicking on a hypertext link, the browser builds an HTTP request and sends it to the IP address indicated by the URL. The HTTP daemon in the destination server machine receives the request and, after any necessary processing, the requested file is returned.
When an HTTP request is processed, it is sent to a particular HTTP server. As described above, various metrics have been implemented for selecting a suitable server on the Internet. However, existing metrics only consider the link between the client's DNS server and the server that the client is attempting to reach. Thus, it would also be desirable if a mechanism for selecting the optimum server (or link) via a preferred service provider could be implemented that takes into account the entire network, from client to server, when making server site selections.
In view of the above, a system and method for redirecting traffic in a distributed system according to individual service provider capabilities would be desirable.
The present invention enables a service request received from a client to be routed to an optimum service provider for processing. This may be accomplished in two different modes: DNS mode and HTTP mode. In either of these modes, the present invention enables a service request to be processed by a type of service (e.g., service provider) that is determined dynamically to be the most efficient and available network for that particular transaction.
In accordance with one embodiment, a network device is configured for performing a method of processing a service request. For example, the service request may be a TCP connection request or a DNS request. When a service request is received, a plurality of packets are sent in response to receiving the service request, where each of the plurality of packets identify a different type of service via which to send the corresponding packet. In addition, the network device maintains a mapping of each different type of service to an IP address. The first one of the plurality of packets to be received may then be used to identify the “winning” type of service (e.g., service provider). Since a mapping of an IP address corresponding to the winning service is maintained, an IP address associated with the winning service may easily be obtained (or provided in the plurality of packets). In this manner, the service request may be processed via an IP address associated with a type of service identified in a first one of the plurality of packets to be received.
In accordance one aspect of the invention, a network device is configured for operating in DNS mode to perform a method of processing a DNS request. First, a DNS request indicating a domain name for which an IP address is requested is received. The network device then transmits a plurality of DNS responses, where each of the plurality of DNS responses is transmitted via a different path associated with a different type of service. Each type of service is mapped to a different IP address. Thus, each of the plurality of DNS responses may include either an IP address associated with the type of service via which the DNS response is transmitted. Alternatively, each of the DNS responses may include a type of service field that identifies the type of service. The DNS responses may then be routed via one or more intermediate routers configured to perform next-hop policy based routing based upon the type of service. The first DNS response to travel through its corresponding path “wins.” For instance, the first DNS response to be received via an intermediate client DNS server that received the initial DNS request from the client is the winning response. In this manner, the optimum route and therefore the fastest, most efficient service provider for that particular transaction may be identified. The client DNS server may then discard all subsequently received DNS responses. The client DNS server may then respond with the winning service to the client. For example, the client DNS server may obtain the IP address of the winning type of service from a table that maps each type of service to a different IP address. In this manner, the client may receive the best IP address for their service type via which to send a TCP packet to initiate a TCP connection.
In accordance with another aspect of the invention, a network device is configured to perform a method of operating in HTTP mode to establish a TCP connection. First, a TCP connection request is received from a client. The network device then sends a plurality of TCP acknowledgement packets to the client via a plurality of paths, each of the plurality of paths corresponding to a different type of service. An acknowledgment message is then received from the client that indicates receipt of one of the plurality of TCP acknowledgement packets sent by the network device. The network device then ascertains the type of service via which the TCP acknowledgement packet received by the client was transmitted. An HTTP redirect is then provided to an IP address corresponding to the type of service.
In the following description, numerous specific details are set forth in order to provide a thorough understanding of the present invention. It will be obvious, however, to one skilled in the art, that the present invention may be practiced without some or all of these specific details. In other instances, well known process steps have not been described in detail in order not to unnecessarily obscure the present invention.
The present invention may operate in two different modes. A DNS mode operates to enable a DNS request to be processed by one of multiple service providers. Similarly, an HTTP mode enables an HTTP request to be processed by an optimum service provider. This is accomplished, in part, through a content router such as that illustrated in
In accordance with the DNS mode, each content router is configured with one or more A-records. An A-record is a record type that is used to translate a domain name into an IP address. In other words, each A-record provides a domain name to IP-address mapping. Thus, in order to assist in a service provider selection, the content router stores IP addresses for multiple service providers.
Once the DNS responses are sent by the content router, intermediate routers perform next-hop policy based routing at block 408. More particularly, each of the intermediate routers is preferably configured to route based upon the TOS field of the DNS responses. For instance, the next-hop routing may be performed based on type of service (TOS) or differentiated services code point (DSCP), which is an extension to TOS. In other words, each DNS response is routed via a different TOS or service provider/facility. The first DNS response to reach the client DNS server “wins,” as shown at block 410. Subsequent DNS responses are discarded by the client DNS server. In this manner, it is possible to determine the fastest and most available network per request. When the client DNS server receives the “winning” DNS response (e.g., A-record), it determines from the A-record the IP address of the winning facility or service and provides this IP address to the requesting client at block 412. More particularly, the IP address may be obtained in variety of ways depending upon the format of the DNS response (e.g., A-record). As one example, if the response includes a different IP address that is mapped to a type of service, service, or service provider, this IP address may be obtained from the DNS response directly. As another example, if the response includes a type of service field, the client DNS server may obtain the IP address mapped to the type of service from a table maintained by the client DNS server. The requesting client sends a TCP packet with its synchronize (SYN) flag set to the winning IP address in order to open a TCP connection at block 414. The web server associated with the service provider mapped to the winning IP address then provides service at block 416.
The HTTP mode leverages the information provided in a TCP packet to enable HTTP requests to be routed to the optimum service provider. As shown in
In accordance with the HTTP mode, an HTTP request is sent by a client at block 602. More particularly, a TCP connection request is sent by the client by sending a TCP packet with its SYN flag set (i.e., TCP syn packet) on port 80. The content router receives the TCP syn packet from the client at block 604. The content router then sends N TCP packets with both the SYN and ACK flags set (i.e., TCP syn ack packets) to the client to attempt to set up a TCP connection through multiple paths as shown at block 606. More particularly, through encapsulating the packet, each of the N TCP packets are sent via a different type of service that is identified within the corresponding TCP packet. This is accomplished as described above with reference to the DNS mode, since the intermediate routers are configured to perform next-hop routing based upon the type of service specified in the TCP packets. In addition, the sequence number in each TCP packet sent by the content router indicates an order that the TCP packets are sent through the different paths. In order to identify a type of service, each TCP packet preferably further identifies a type of service. This may be accomplished by indicating the type of service in the reserved field as described above with reference to block 510 of
Generally, the techniques of the present invention may be implemented on software and/or hardware. For example, they can be implemented in an operating system kernel, in a separate user process, in a library package bound into network applications, on a specially constructed machine, or on a network interface card. In a specific embodiment of this invention, the technique of the present invention is implemented in software such as an operating system or in an application running on an operating system.
A software or software/hardware hybrid implementation of the techniques of this invention may be implemented on a general-purpose programmable machine selectively activated or reconfigured by a computer program stored in memory. Such programmable machine may be a network device designed to handle network traffic, such as, for example, a router or a switch. Such network devices may have multiple network interfaces including frame relay and ISDN interfaces, for example. Specific examples of such network devices include routers and switches. For example, the content router, DNS server, and intermediate routers of this invention may be specially configured routers or servers such as specially configured router models 1600, 2500, 2600, 3600, 4500, 4700, 7200, 7500, and 12000 available from Cisco Systems, Inc. of San Jose, Calif. A general architecture for some of these machines will appear from the description given below. In an alternative embodiment, the techniques of this invention may be implemented on a general-purpose network host machine such as a personal computer or workstation. Further, the invention may be at least partially implemented on a card (e.g., an interface card) for a network device or a general-purpose computing device.
Referring now to
CPU 862 may include one or more processors 863 such as a processor from the Motorola family of microprocessors or the MIPS family of microprocessors. In an alternative embodiment, processor 863 is specially designed hardware for controlling the operations of network device 860. In a specific embodiment, a memory 861 (such as non-volatile RAM and/or ROM) also forms part of CPU 862. However, there are many different ways in which memory could be coupled to the system. Memory block 861 may be used for a variety of purposes such as, for example, caching and/or storing data, programming instructions, etc.
The interfaces 868 are typically provided as interface cards (sometimes referred to as “line cards”). Generally, they control the sending and receiving of data packets over the network and sometimes support other peripherals used with the network device 860. Among the interfaces that may be provided are Ethernet interfaces, frame relay interfaces, cable interfaces, DSL interfaces, token ring interfaces, and the like. In addition, various very high-speed interfaces may be provided such as fast Ethernet interfaces, Gigabit Ethernet interfaces, ATM interfaces, HSSI interfaces, POS interfaces, FDDI interfaces and the like. Generally, these interfaces may include ports appropriate for communication with the appropriate media. In some cases, they may also include an independent processor and, in some instances, volatile RAM. The independent processors may control such communications intensive tasks as packet switching, media control and management. By providing separate processors for the communications intensive tasks, these interfaces allow the master microprocessor 862 to efficiently perform routing computations, network diagnostics, security functions, etc.
Although the system shown in
Regardless of network device's configuration, it may employ one or more memories or memory modules (such as, for example, memory block 865) configured to store data, program instructions for the general-purpose network operations and/or other information relating to the functionality of the techniques described herein. The program instructions may control the operation of an operating system and/or one or more applications, for example. The memory or memories may also be configured to include a mapping of IP addresses to an associated type of service, as well as a plurality of A-records for use by a content router in the DNS mode, etc.
Because such information and program instructions may be employed to implement the systems/methods described herein, the present invention relates to machine readable media and computer-readable medium that include program instructions, state information, etc. for performing various operations described herein. Examples of machine-readable media include, but are not limited to, magnetic media such as hard disks, floppy disks, and magnetic tape; optical media such as CD-ROM disks; magneto-optical media such as floptical disks; and hardware devices that are specially configured to store and perform program instructions, such as read-only memory devices (ROM) and random access memory (RAM). The invention may also be embodied in a carrier wave travelling over an appropriate transmission medium such as airwaves, optical lines, electric Lines, etc. Examples of program instructions include both machine code, such as produced by a compiler, and files containing higher level code that may be executed by the computer using an interpreter.
Although illustrative embodiments and applications of this invention are shown and described herein, many variations and modifications are possible which remain within the concept, scope, and spirit of the invention, and these variations would become clear to those of ordinary skill in the art after perusal of this application. For instance, the present invention is described as being implemented and capable of operating in DNS mode or HTTP redirect mode. However, it should be understood that the invention is not limited to such implementations, but instead would equally apply regardless of the context and system in which it is implemented. Thus, broadly speaking, the operations described above may be used to distribute traffic in a distributed system based on a type of service (e.g., service, service provider, network connection or domain) that is determined to provide the optimum and most efficient path. Thus, although the service request is described as a TCP connection request or a DNS request, the service request may be implemented using other connection initiation protocols as well. In addition, the above-described invention may be stored on a disk drive, a hard drive, a floppy disk, a server computer, or a remotely networked computer. Accordingly, the present embodiments are to be considered as illustrative and not restrictive, and the invention is not to be limited to the details given herein, but may be modified within the scope and equivalents of the appended claims.
Number | Name | Date | Kind |
---|---|---|---|
4827411 | Arrowood et al. | May 1989 | A |
4965772 | Daniel et al. | Oct 1990 | A |
5414704 | Spinney | May 1995 | A |
5452447 | Nelson et al. | Sep 1995 | A |
5488412 | Majeti et al. | Jan 1996 | A |
5506987 | Abramson et al. | Apr 1996 | A |
5511208 | Boyles et al. | Apr 1996 | A |
5555244 | Gupta et al. | Sep 1996 | A |
5586121 | Moura et al. | Dec 1996 | A |
5611049 | Pitts | Mar 1997 | A |
5673265 | Gupta et al. | Sep 1997 | A |
RE35774 | Moura et al. | Apr 1998 | E |
5751971 | Dobbins et al. | May 1998 | A |
5774660 | Brendel et al. | Jun 1998 | A |
5787470 | DiSimone et al. | Jul 1998 | A |
5818845 | Moura et al. | Oct 1998 | A |
5828655 | Moura et al. | Oct 1998 | A |
5848241 | Misinai et al. | Dec 1998 | A |
5852717 | Bhide et al. | Dec 1998 | A |
5859852 | Moura et al. | Jan 1999 | A |
5872773 | Katzela et al. | Feb 1999 | A |
5892903 | Klaus | Apr 1999 | A |
5946047 | Levan | Aug 1999 | A |
5946048 | Levan | Aug 1999 | A |
5950205 | Aviani, Jr. | Sep 1999 | A |
5953335 | Erimli et al. | Sep 1999 | A |
5956346 | Levan | Sep 1999 | A |
5959660 | Levan | Sep 1999 | A |
5959968 | Chin et al. | Sep 1999 | A |
5959997 | Moura et al. | Sep 1999 | A |
5989060 | Coile et al. | Nov 1999 | A |
6006264 | Colby et al. | Dec 1999 | A |
6006266 | Murphy, Jr. et al. | Dec 1999 | A |
6016388 | Dillon | Jan 2000 | A |
6052718 | Gifford | Apr 2000 | A |
6182139 | Brendel | Jan 2001 | B1 |
6272550 | Deng et al. | Aug 2001 | B1 |
6298380 | Coile et al. | Oct 2001 | B1 |
6304913 | Rune | Oct 2001 | B1 |
6345294 | O'Toole et al. | Feb 2002 | B1 |
6411986 | Susai et al. | Jun 2002 | B1 |
6430623 | Alkhatib | Aug 2002 | B1 |
6587438 | Brendel | Jul 2003 | B1 |
6789125 | Aviani et al. | Sep 2004 | B1 |
6810411 | Coughlin et al. | Oct 2004 | B1 |
6810421 | Ishizaki et al. | Oct 2004 | B1 |
6950429 | Benayoun et al. | Sep 2005 | B2 |
7103651 | Bohannon et al. | Sep 2006 | B2 |
20020112036 | Bohannon et al. | Aug 2002 | A1 |
20050172011 | Gourlay et al. | Aug 2005 | A1 |
Number | Date | Country |
---|---|---|
WO 9831107 | Jul 1998 | WO |