摘要:
A method, computer program product and system for processing TCP/IP packets. A TCP protocol stack may store a payload of a received TCP/IP packet in a data fragment list. The TCP protocol stack may further read the header of the received packet to extract a value used to index into a table storing a list of transport control blocks (TCBs). The TCP protocol stack may further perform a lock and a read operation on the TCB indexed in the table. The TCP protocol stack may further transmit the payload to the TCP application without requiring the application to perform a lock, read, write or unlock operation on the indexed TCB since the TCP protocol stack and the TCP application are operating on the same thread. By the TCP application foregoing the lock, read, write and unlock operations on the TCB, there is a reduction in the number of memory accesses.
摘要:
A method, computer program product and system for processing TCP/IP packets. A TCP protocol stack may store a payload of a received TCP/IP packet in a data fragment list. The TCP protocol stack may further read the header of the received packet to extract a value used to index into a table storing a list of transport control blocks (TCBs). The TCP protocol stack may further perform a lock and a read operation on the TCB indexed in the table. The TCP protocol stack may further transmit the payload to the TCP application without requiring the application to perform a lock, read, write or unlock operation on the indexed TCB since the TCP protocol stack and the TCP application are operating on the same thread. By the TCP application foregoing the lock, read, write and unlock operations on the TCB, there is a reduction in the number of memory accesses.
摘要:
Method for extending the crankback procedure in an Asynchronous Transfer Mode (ATM) system composed of at least a first data network (10) comprising a plurality of switching nodes interconnected by connection lines and including end switching nodes each being connected to at least a Data Transmission equipment (DTE) and being used either as an entry border node (22) when it is connected to a source DTE (18) or an exit border node (28) when it is connected to a destination DTE (20), the network using a routing protocol of the type wherein a best route between a source DTE and a destination DTE is determined in a control point associated with the entry border node to which is connected the source DTE and wherein a set-up message is sent by the entry border node, and a second data network (12) including at least one DTE to be used as destination DTE in an exchange of data with a source DTE connected to the first data network and being interconnected with the first data network by means of at least two links (14, 16) not supporting the routing protocol. Such a method consists, when the exit border node of the first data network receives a clearing message on one link indicating that the set-up message has been rejected because the best route is blocked anywhere in the second data network, in building a crankback information element to be added to the clearing message for enabling the entry border node to find an alternate route avoiding the portion of the route which is blocked.
摘要:
A method, computer program product and system for reducing the number of messages to be processed by a control processor in a load balancer. A network processor instead of the control processor in the load balancer may establish and terminate a TCP connection between a client and the load balancer and between the load balancer and a server. Further, messages, e.g., data and control messages, may be bundled into a single message to be transmitted between the network processor and the control processor. By bundling messages into a single message as well as offloading the establishment and termination of TCP connections to the network processor instead of the control processor, the number of messages to be processed by the control processor is reduced. Consequently, the performance of the load balancer is improved.
摘要:
A system for providing a scalable processor and operating system independent network processor services architecture. The system includes a plurality of portable and individualized functional components representing particular segments of the control processor's device driver. The functional components, which include lower level and external APIs, carry out the various network processor functions such as the receipt and transfer of packets on the network, and other functions required by the control processor to communicate with and direct the network processor. The functional components are designed to be adaptable to the various types of processor architecture and operating systems available and to permit customers or developers to customize and expand the available network services.
摘要:
A method, system and article of manufacture to prevent signaling congestion in a connection-oriented network node in situations where a plurality of network-attached data terminal equipments (source DTEs) concurrently request a connection to at least one network-attached data terminal equipment (destination DTE), each of the source DTEs sending call setup messages (CSMs) through the network node to the at least one destination DTE. The CSMs having to be processed by the network to establish the connections requested.
摘要:
A method and system for performing network address translations for a session in a network is disclosed. The network includes at least one local network domain, and the at least one local network domain includes at least one computer system. Each computer system has a local address and is associated with a global address. The session exchanges packets that travel between the computer system within the local network domain and another computer system which may be outside of the local network domain. Each packet includes source and destination information. The method and system include determining a direction of travel for each packet, inbound or outbound, by searching a global address table for a match of a key for each packet. The key is provided using a portion of the destination information. The global address table includes at least one entry. Each entry corresponds to the global address for a first corresponding computer system. The method and system also include asymmetrically translating the source and destination information for each packet using an address translation table or session table based on whether or not the full match is found. The destination information is translated using information in the session table if the packet is inbound. The source information is translated based on the address translation table if the packet is outbound. The address translation table includes at least one entry. Each entry corresponds to the local address for a first corresponding computer system or a global host name for a shared host.
摘要:
A timer management system and method for managing timers in both a synchronous and asynchronous system. In one embodiment of the present invention, a timer management system comprises an application program interface (API) for providing a set of synchronous functions allowing an application to functionally operate on the timer. The timer management system further comprises a timer database for storing timer-related information. Furthermore, the timer management system comprises a timer services for detecting the expiring of the timer. A handle function of the timer services allows an asynchronous application, i.e., application in a multi-task system, to synchronously act on the timer. That is, when a timer in a asynchronous system times-out, the handle function allows the asynchronous application to act on the expired timer without incurring an illegal time-out message. In another embodiment of the present invention, a timer may be reinitialized from the same allocated block of memory used to create the timer. In another embodiment of the present invention, a time-out message may be sent using the same allocated block of memory used to create the timer.