Abstract: When we launched the Akamai system in early 1999, it initially delivered only Web objects (images and documents). It has since evolved to distribute dynamically generated pages and even applications to the network's edge, providing customers with on-demand bandwidth and computing capacity. This reduces content providers' infrastructure requirements, and lets them deploy or expand services more quickly and easily. Our current system has more than 12,000 servers in over 1,000 networks. Operating servers in many locations poses many technical challenges, including how to direct user requests to appropriate servers, how to handle failures, how to monitor and control the servers, and how to update software across the system. We describe our system and how we've managed these challenges.

Abstract: Methods for load rebalancing by clients in a network are disclosed. Client load rebalancing allows the clients to optimize throughput between themselves and the resources accessed by the nodes. A network which implements this embodiment of the invention can dynamically rebalance itself to optimize throughput by migrating client I/O requests from overutilized pathways to underutilized pathways. Client load rebalancing refers to the ability of a client enabled with processes in accordance with the current invention to remap a path through a plurality of nodes to a resource. The remapping may take place in response to a redirection command emanating from an overloaded node, e.g. server. These embodiments disclosed allow more efficient, robust communication between a plurality of clients and a plurality of resources via a plurality of nodes. In an embodiment of the invention a method for load balancing on a network is disclosed. The network includes at least one client node coupled to a plurality of server nodes, and at least one resource coupled to at least a first and a second server node of the plurality of server nodes. The method comprises the acts of: receiving at a first server node among the plurality of server nodes a request for the at least one resource; determining a utilization condition of the first server node; and re-directing subsequent requests for the at least one resource to a second server node among the plurality of server nodes in response to the determining act. In another embodiment of the invention the method comprises the acts of: sending an I/O request from the at least one client to the first server node for the at least one resource; determining an I/O failure of the first server node; and re-directing subsequent requests from the at least one client for the at least one resource to an other among the plurality of server nodes in response to the determining act.

Abstract: The overall increase in traffic on the World Wide Web is augmenting user-perceived response times from popular Web sites, especially in conjunction with special events. System platforms that do not replicate information content cannot provide the needed scalability to handle large traffic volumes and to match rapid and dramatic changes in the number of clients. The need to improve the performance of Web-based services has produced a variety of novel content delivery architectures. This article will focus on Web system architectures that consist of multiple server nodes distributed on a local area, with one or more mechanisms to spread client requests among the nodes. After years of continual proposals of new system solutions, routing mechanisms, and policies (the first dated back to 1994 when the NCSA Web site had to face the first million of requests per day), many problems concerning multiple server architectures for Web sites have been solved. Other issues remain to be addressed, especially at the network application layer, but the main techniques and methodologies for building scalable Web content delivery architectures placed in a single location are settled now. This article classifies and describes main mechanisms to split the traffic load among the server nodes, discussing both the alternative architectures and the load sharing policies. To this purpose, it focuses on architectures, internal routing mechanisms, and dispatching request algorithms for designing and implementing scalable Web-server systems under the control of one content provider. It identifies also some of the open research issues associated with the use of distributed systems for highly accessed Web sites.

Abstract: A management tool that streamlines the server allocation and provisioning processes within a data center is provided. The system, method, and computer program product divide the server provisioning and allocation into two separate tasks. Provisioning a server is accomplished by generating a fully configured, bootable system image, complete with network address assignments, virtual LAN (VLAN) configuration, load balancing configuration, and the like. System images are stored in a storage repository and are accessible to more than one server. Allocation is accomplished using a switching mechanism which matches each server with an appropriate system image based upon current configuration or requirements of the data center. Thus, real-time provisioning and allocation of servers in the form of automated responses to changing conditions within the data center is possible. The ability to instantly re-provision servers, safely and securely switch under-utilized server capacity to more productive tasks, and improve server utilization is also provided.

Abstract: An address of a server (406) that should supply an information object or service to a requestor is returned in response to a request therefor. The address (416) of the server (406) that is returned is an optimal server (406) selected according to specified performance metrics. The specified performance metrics may include one or more of an average delay from the server to another, average processing delays at the server, reliability of a path from the server to another, available bandwidth in said path, and loads on the server.

Abstract: A user-centric session management system and method are provided, in which a user remains authenticated and connected to a session while migrating between provisioning applications, protocols and/or client devices. Each user has a unique user identification (UI), and each session has a unique session identifier (USI). The USI supports anonymous users and maintains authentication without requiring authentication for each request. The system includes a session manager that accesses session state memory and virtual device memory. The session state memory provides short-term storage of records of all current client-server sessions, including USIs and associated UIs). The virtual device memory provides long-term storage of state mirroring the current state of a client device involved in a transaction during a session. Using the USI or UI and the virtual device associated therewith, the client device is synchronized at re-connect to an ongoing session or to an interrupted transaction associated with a terminated session.

Abstract: The invention provides a session-based bi-directional multi-tier client-server asynchronous information database search and retrieval system for sending a character-by-character string of data to an intelligent server that can be configured to immediately analyze the lengthening string character-by-character and return to the client increasingly appropriate database information as the client sends the string.

Abstract: Popular content is frequently replicated in multiple servers or caches in the Internet to offload origin servers and improve end-user experience. However, choosing the best server is a nontrivial task and a bad choice may provide poor end user experience. In contrast to retrieving a file from a single server, we propose a parallel-access scheme where end users access multiple servers at the same time, fetching different portions of that file from different servers and reassembling them locally. The amount of data retrieved from a particular server depends on the resources available at that server or along the path from the user to the server. Faster servers will deliver bigger portions of a file while slower servers will deliver smaller portions. If the available resources at a server or along the path change during the download of a file, a dynamic parallel access will automatically shift the load from congested locations to less loaded parts (server and links) of the Internet. The end result is that users experience significant speedups and very consistent response times. Moreover, there is no need for complicated server selection algorithms and load is dynamically shared among all servers. The dynamic parallel-access scheme presented in this paper does not require any modifications to servers or content and can be easily included in browsers, peer-to-peer applications or content distribution networks to speed up delivery of popular content.

Abstract: In a site-sensitive Internet service system enhanced client browsers report URLs to an Internet-connected service-control server, which informs clients of other clients visiting common URLs, and organizes and enables chat sessions among simultaneous visitors to common Internet sites. In some cases the server activity, including chat sessions is transparent to hosts of the WEB sites visited. In others the WEB sites, through cooperation with the service-control server, control information streaming to clients, including chat sessions. Information streamed may include banner ads enabled as hyperlinks. In a further embodiment the service-control server maintains and reports statistics based on site visitation reported by clients.

Abstract: A system for permitting passengers on board an aircraft to send and receive electronic data is described. The components of the system on board the aircraft include a server having a plurality of nodes to which computer terminals are attached, as desired. The components of the system on board the aircraft include a wireless access point having a plurality nodes, where the wireless access point is attached to the server and to a plurality of wireless cards inserted into computer terminals, as desired. The computer terminals are laptop or palm-top personal computers belonging to the various passengers on board or fixed terminals within the aircraft. The server communicates with a wide variety of different terminals running different operating systems and with the access point. Each computer terminal is connected to the server via an aircraft cable or wireless network. Server has mass storage which contains a database of WWW pages which can be browsed by passengers using terminals. Server provides a domain name server (DNS) that masquerades as the passenger's usual DNS. Server then links the passenger to the appropriate locally stored WWW page. Server also contains storage for e-mail messages. Connected to server is one or more radios. This permits data to be transferred to base station using communications network. A virtual private network (VPN) connects station to communications service provider networks, web content processor, and via links to the Internet, including access to subscriber ISPs/corporate mail servers and other mail servers. Points of Presence (POP) provide Internet access and e-mail service to subscribers of the service while not on the aircraft. POPs can also be used by communications service provider networks and web content processors as an alternate means to connect to VPN.

Abstract: A content-aware flow switch intercepts a client content request in an IP network, and transparently directs the content request to a best-fit server. The best-fit server is chosen based on the type of content requested, the quality of service requirements implied by the content request, the degree of load on available servers, network congestion information, and the proximity of the client to available servers. The flow switch detects client-server flows based on the arrival of TCP SYNs and/or HTTP GETs from the client. The flow switch implicitly deduces the quality of service requirements of a flow based on the content of the flow. The flow switch also provides the functionality of multiple physical web servers on a single web server in a way that is transparent to the client, through the use of virtual web hosts and flow pipes.

Abstract: A system and method is provided for handling network communications between a client and a target server on the Internet to protect the privacy and anonymity of the client. For a session between the client and the target server, a routing control server sets up a routing chain using a plurality of Web servers randomly selected from a pool of participating Web servers as routers for routing messages between the client and the target server. To prevent traffic analysis, an “onion encryption” scheme is applied to the messages as they are forwarded along the routing chain. A payment service cooperating with the routing control server allows a user to pay for the privacy protection service without revealing her real identity.

Abstract: A single-sign-on process and mechanism for a client who wishes to access multiple servers in an environment, where the servers employ the Kerberos authentification process. During an initial log in process to a first server by the client, the first server performs a Kerberos authentification on the client and stores the ticket-granting ticket (TGT) for that client in server memory. The first server then provides the client with a token corresponding to that stored TGT, but does not transmit the TGT itself to the client. When the client requests service from subsequent server, the client provides the token with the request. The subsequent server then requests the client's TGT from the first server using the client-supplied token. The first server retrieves the TGT from memory, and transmits it to the subsequent server. The subsequent server then may use the TGT to determine if the client is authorized to access the service or resource requested.

Abstract: The present invention provides systems and methods for maintaining stateful interactions between clients and servers. Furthermore, the invention provides systems and methods for maintaining stateful interactions between clients and load balancers. In one embodiment, the present invention provides systems and methods for maintaining statefulness without the need for the server to query and/or store information on the client.

Abstract: A content-aware application switch and methods thereof intelligently switch client packets to one server among a group of servers in a server farm. The switch uses Layer 7 or application content parsed from a packet to help select the server and to schedule the transmitting of the packet to the server. This enables refined load-balancing and Quality-of-Service control tailored to the application being switched. In another aspect of the invention, a slow-start server selection method assigned an initially boosted server load metric to a server newly added to the group of servers under load balancing. This alleviates the problem of the new server being swamped initially due to a very low load metric compared to that of others. In yet another aspect of the invention, a switching method dependent on Layer 7 content avoids delayed binding in a new TCP session. Layer 7 content is not available during the initial handshaking phase of a new TCP session. The method uses the Layer 7 content from a previous session as an estimate to help select the server and uses a default priority to scheduling the transmitting of the handshaking packets. Updated Layer 7 content available after the handshaking phase is then used to reset the priority for the transmit schedule and becomes available for use in load balancing of the next TCP session.

Abstract: In one aspect, a mapping between a file type and a program is received, a client system file is graphically presented, a selection of the graphical depiction is received, a program associated with the selected file is identified, and a request to execute the identified program is sent. In another aspect, a server system file is graphically presented, a selection of the graphical depiction is received, the selection is transmitted to a server system, a request to execute an associated program is received, and the associated program is executed. In still another aspect, a web server file is graphically presented and a request to execute a first program that includes information concerning a second program associated with the file is received. Application output data from the second program is reviewed and formatted by the first program and received application output data is presented to the user.

Abstract: A system and a method for transparent takeover (or failover) of a remote client TCP connection from a first server in a cluster of interconnected servers to a second server provides the storing of shared state information relative to the connection on each involved server and using of the shared state information to reestablish the connection on the second server. A message using a sequence number to refer to a previously transmitted data element (such as a byte) is sent by the second server and a received client acknowledgement (ACK) of that sequence number, or a higher one, is used to synchronize the server's data packet transmission sequence number with the ACK-transmitted sequence number. If synchronization is successful, then the connection is restarted on the second server from the point of termination/failure on the first server.

Abstract: A method and system for sharing content among client computer systems. A content sharing system includes a server component executing on a server computer system that controls the logging on and off of users and establishes a connection or session with each client computer system whose user is logged-on. Because of these connections, administrative messages from one client to any other client can be sent via the server. The shared content, however, may be sent from client to client on a peer-to-peer basis without sending the shared content to the server.

Abstract: A peer-to-peer network has a server that maintains a list of addresses of clients connected to the peer-to-peer network. The server generates seed lists for each of the clients connected to the network from the server maintained list. The clients conduct peer-to-peer searches using the seed list provided by the server. The seed lists may be either randomly generated or generated to facilitate network objectives. The clients may periodically request seed list refreshing. The peer-to-peer network may further include authentication and authorization search security checks.

Abstract: A apparatus and method for effecting a distributed video on demand system is a local data network. A number of local processing units within the local network are utilized, with the local processing units being associated with system subscribers of the video on demand system. The local processing units include memories that are each configured to store video data. A database server in communication with local data network is further included for directing streamed delivery of the video data from certain ones of the local processing units storing particular video data to other local processing units in the local network that are requesting playback of the particular video data.

Abstract: A system for permitting passengers on board an aircraft to send and receive electronic data is described. The components of the system on board the aircraft include a server having a plurality of nodes to which computer terminals are attached, as desired. The components of the system on board the aircraft include a wireless access point having a plurality nodes, where the wireless access point is attached to the server and to a plurality of wireless cards inserted into computer terminals, as desired. The computer terminals are laptop or palm-top personal computers belonging to the various passengers on board or fixed terminals within the aircraft. The server communicates with a wide variety of different terminals running different operating systems and with the access point. Each computer terminal is connected to the server via an aircraft cable or wireless network. Server has mass storage which contains a database of WWW pages which can be browsed by passengers using terminals. Server provides a domain name server (DNS) that masquerades as the passenger's usual DNS. Server then links the passenger to the appropriate locally stored WWW page. Server also contains storage for e-mail messages. Connected to server is one or more radios. This permits data to be transferred to base station using communications network. A virtual private network (VPN) connects station to communications service provider networks, web content processor, and via links to the Internet, including access to subscriber ISPs/corporate mail servers and other mail servers. Points of Presence (POP) provide Internet access and e-mail service to subscribers of the service while not on the aircraft. POPs can also be used by communications service provider networks and web content processors as an alternate means to connect to VPN.

Abstract: A system and method are shown for load balancing across global network resources using an existing network protocol, such as Mobile IP, having a redirect feature. According to one method, each of a plurality of servers at a data center uses Mobile IP to obtain an IP address that is also provided to a content server site. Further, a content server site includes a plurality of IP addresses assigned to the plurality of servers and creates a load database including load data for each server. When a client request is received at the content server site from a client device, the content server site determines a network address of a server to process the client request based on the load data, and provides the network address of the server to the client device. When the client device receives the network address of the server, the client device sends an application request to the selected server, and the selected server sends an application response to the client device.

Abstract: Mechanisms and techniques provide a method for determining the performance of a server, based upon a connection characteristic of a connection between the server and a client. The connection characteristic, such as the window size of the server, provides information regarding the performance or capacity of the server. Variation between the connection characteristic and a preferred connection characteristic is used by a data communications device, for example, to determine the performance of the server. Based upon the variation in the connection characteristic, the data communications device is then able to direct incoming content requests to other servers in communication with a network, as required.

Abstract: A system and method for managing a distributed data processing network are provided. A distributed network environment is configured such that monitoring and control devices are associated with device servers in a secure subnet. Each device server connects with a premises server. According to the present invention, a client computing device utilizing a WWW browser employs a communication protocol to pass commands to device servers and devices through the premises server. In another aspect of the present invention, a distributed computing environment allows multiple device servers to cumulatively process data collected from cameras, sensors, and other attached devices and provide a common computing platform and user interface.

Abstract: A server stores files. Distributed clients access the server, to learn about changes made to the files on the server, and to push local changes of the files onto the server. A synchronization application is used to synchronize the clients and server, synchronizing metadata and selected files.

Abstract: A network-based system is provided for enabling agent-users of the system to obtain current client-status information related to clients of an information-source facility connected to the network in order to optimize callback-connection success from the agent-users to the monitored clients. The system comprises, a first server node connected to the information-source facility and to the network, a second server node connected to the first server node and to the network, the second server node accessible to the first server node, a network-capable appliance connected to the network, the second server node accessible to the network-capable appliance, an agent workstation connected to the network and to the first server node, the first server node accessible to the agent workstation and a software application distributed on at least the first and second server nodes, the software application enabling distribution of the client-status information. The agent-user operating the agent workstation accesses the first server node and subscribes to the client-status information, the client-status information is accessed from the second server node by the first server node and delivered to the requesting agent-user.

Abstract: A system and method for enabling failover in an application server cluster. A “primary” application server computer in the cluster may provide a service or data necessary for other application server computers in the cluster to operate. In addition to the primary application server computer, one or more of the other application server computers may be designated as “backup” application server computers. Each backup application server may backup the processing information managed by the primary application server. When the primary application server itself becomes unavailable (e.g., due to a failure of the computer system or network), one or more of the backup application servers may be promoted to the role of primary application server.

Abstract: Method and system for a service server to provide a service to a client. The client (C) sets up a secure session to an authentication server (CAP) and sends its identifier and a service request stating the required service. The authentication server verifies the client identifier and sends the service request to a service authorization server (DAP). The authorization server checks whether the required service may be provided and sends the authorized service request to the authentication server. The authentication server generates a token, associated with the authorized service request. Via the secure session, the authentication server sends the address of the relevant service server and the token. The client sends the token to the service server, which then sends the token to the authentication server. The authentication server fetches the service request associated with the token and forwards it to the service server, after which the service server gives the client the required service.

Abstract: A virtualized logical server cloud (201) that enables logical servers (106) to exist independent of physical servers (108) that instantiate the logical servers (106). Servers are treated as logical resources in order to create a logical server cloud. The logical attributes of a logical server (106) are non-deterministically allocated to physical resources creating a cloud of logical servers (106) over physical servers (108). Logical separation is facilitated by the addition of a server cloud manager (102). Each physical server (108) includes or is coupled to physical resources including a network resource (217), a data storage resource (219) and a processor resource (225). At least one physical server executes virtualization software (229) that virtualizes physical resources for logical servers (106). The server cloud manager 102 maintains status and instance information for the logical servers (106) including persistent and non-persistent attributes that link each logical server (106) with a physical server (108).