International Business Machines Corporation

Satya Prakesh Sharma

Vinit Jain

James P Allen

Deanna Lynn Quigg Brown

Herman Dietrich Dierks

Tai-Chien Daisy Chang

Shawn Patrick Mullen

Giles R Frazier

Daniel Gene Eisenhauer

Venkat Venkatsubra

Rakesh Sharma

Jeffrey Paul Messing

Robert G Kovacs

Gregory Scott Althaus

William Christopher Conklin

William James Hymas

Helmut Cossmann

Methods, systems, and computer program products are provided for migrating an operating system from a source computer to a destination computer. Some embodiments include identifying a destination adapter of the destination computer that is coupled for data communications to a SAN that is also coupled for data communications to a source adapter of the source computer, logging off a virtual port name of the source adapter from the login manager, deregistering the virtual port name from the source adapter, registering the virtual port name with the destination adapter, and logging on the destination adapter to the login manager with the virtual port name. Typical embodiments also include transferring the operating system from the source computer to the destination computer.

Operating system migration with minimal storage area network reconfiguration

A method, system and apparatus for negotiating parameters for an IPSec connection between a requesting client and an iSCSI system using a computer system other than an iSCSI system are provided. By design, the iSCSI system monitors TCP (Transmission Control protocol) portfor secure requests. When a request enters port, the iSCSI system transmits all information received on portto a computer system better suited to handle IPSec parameter negotiations. After the computer system has negotiated the parameters, the parameters are passed to the iSCSI system for a secure data transaction to ensue.

ISCSI target offload administrator

A method, system and apparatus for multicasting or broadcasting a piece of data from one partition of a partitioned computer system to a plurality of partitions of the system are provided. When a partition needs to transfer data to more than one partition of the system, the partition first stores the data into a global memory accessible by all partitions of the system. The memory is then marked &ldquo;read-only&rdquo; to ensure that the data is not over-written and the Ids (identifications) of the recipient partitions are passed to a piece of firmware or hardware of the computer system. This firmware or hardware then alerts the recipient partitions that there is a piece of data ready to be read in the memory. Once all the recipient partitions have read the data, the global memory is reverted to a &ldquo;read&rdquo; and &ldquo;write&rdquo; memory.

Apparatus and method of multicasting or broadcasting data from one partition of a partitioned computer system to a plurality of other partitions

A method, computer program product and system for detecting a first-hop dead gateway. In one embodiment, a method comprises the step of sending a TCP packet of data from an application of a sender host to a receiver host through a first gateway, where the first gateway is a first-hop away from the sender host. The method further comprises the step of TCP failing to receive an acknowledgment of received data from the receiver host. The method further comprises the step of deleting an ARP entry associated with the first gateway in the sender host. The method further comprises the step of establishing a new communication using the first gateway by the application or new application of the sender host. The method further comprises the step of sending an ARP request to the first gateway by the sender host. If the sender host receives a response from the first gateway, then the method further comprises the step of sending a TCP or non-TCP packet of data using the first gateway if the new communication was a TCP or non-TCP communication, respectively. If the sender host does not receive a response from the first gateway, then the method further comprises the step of selecting an alternative path through an alternative first-hop gateway in the routing table of the sender host. The application or new application of the sender host then sends the TCP or non-TCP packet of data using the alternative gateway if the new communication was a TCP or non-TCP communication, respectively.

Detecting a dead gateway for subsequent non-TCP transmission by sending a first TCP packet and deleting an ARP entry associated with the gateway

Network input processing is distributed to multiple CPUs on multiprocessor systems to improve network throughput and take advantage of MP scalability. Packets are received by the network adapter and are distributed to N receive buffer pools set up by the device driver, based on N CPUs being available for input processing of packets. Each receive buffer pool has an associated CPU. Packets are direct memory accessed to one of the N receive buffer pools by using a hashing function, which is based on the source MAC address, source IP address, or the packet's source and destination TCP port numbers, or all or a combination of the foregoing. The hashing mechanism ensures that the sequence of packets within a given communication session will be preserved. Distribution is effected by the network adapter, which sends an interrupt to the CPU corresponding to the receive buffer pool, subsequent to the packet being DMAed into the buffer pool. This optimizes the efficiency of the MP system by eliminating any reliance on the scheduler and increasing the bandwidth between the device driver and the network adapter, while maintaining proper packet sequences. Parallelism is thereby increased on network I/O processing, eliminating CPU bottleneck for high speed network I/Os and, thus, improving network performance.

Network adapter utilizing a hashing function for distributing packets to multiple processors for parallel processing

Network input processing is distributed to multiple CPUs on multiprocessor systems to improve network throughput and take advantage of MP scalability. Packets received on the network are distributed to N high priority threads, wherein N is the number of CPUs on the system. N queues are provided to which the incoming packets are distributed. When one of the queues is started, one of the threads is scheduled to process packets on this queue at any one of the CPUs that is availableat the time. When all of the packets on the queue are processed, the thread becomes dormant. Packets are distributed to one of the N queues by using a hashing function based on the source MAC address, source IP address, or the packet's source and destination TCP port number, or all or a combination of the foregoing. The hashing mechanism ensures that the sequence of packets within a given communication session will be preserved. Distribution is effected by the device drivers of the system. Parallelism is thereby increased on network I/O processing, eliminating CPU bottleneck for high speed network I/Os, thereby improving network performance

System and method for sequencing packets for multiprocessor parallelization in a computer network system

Growing community of inventors

Round Rock, TX, United States of America

Satya Prakesh Sharma