International Business Machines Corporation

John Andrew Aiken, Jr

David Anthony Herr

Mark William McClintock

Wesley McMillan Devine

Alfred Bundgaard Christensen

Michael E Baskey

Sue L Huang

Harry M Yudenfriend

Donald W Schmidt

Jeffrey Mark Nick

Jeffrey Paul Kubala

Jerry Wayne Stevens

Steven G Glassen

Janet R Easton

Frank William Brice

Gary Stuart Puchkoff

Dilip Dinkar Kandlur

Robert Curt Nielsen

Ambrose A Verdibello, Jr

Kenneth O Shipp, Jr

Andrew Henry Richter

James L Hall

Wolfgang Eckert

Raymond Edward Ward

Heinrich K Lindner

Aubrey M Herzik

Marcus Eder

Methods, systems and computer program products provide for establishing a connection originated by an application executing on a data processing system in a cluster of data processing systems. A dynamic network address is associated with the application at the data processing system on which the application is executing. If a request is received for the data processing system to originate a connection that is associated with the application, the connection is established utilizing the dynamic network address associated with the application as a source address for the connection.

Methods, systems and computer program products for source address selection

Connections from a client to respective ones of a plurality of application instances that receive data from a shared communication protocol stack are assigned by defining an affinity between the client and a corresponding one of the plurality of application instances. Connection requests received from the client at the shared communication protocol stack are assigned to the corresponding one of the plurality of application instances if an affinity is defined between the client and the corresponding one of the plurality of application instances. Connection requests received from the client at the shared communication protocol stack are distributed to selected ones of the plurality of application instances if an affinity is not defined between the client and the corresponding one of the plurality of application instances.

Application instance level workload distribution affinities

Methods, systems and computer program products provide communications with a plurality of application instances using a plurality of communication protocol stacks and a single Internet Protocol (IP) address. A stack is established as a routing stack associated with the single IP address. Other stacks which are associated with the single IP address are defined as candidate target stacks and an identification of candidate target stacks and the routing stack is distributed to the stacks. The routing stack is notified when an application associated with a candidate target stack listens to a port of the IP address to establish a current actual target stack. A request to establish a connection to the IP address and the port of the IP address is received and a routing table entry corresponding to the current actual target stack is created to provide a routing path from the routing stack to the current actual target stack. Communications to the port of the IP address received by the routing stack are routed based on the routing table.

System having a single IP address associated with communication protocol stacks in a cluster of processing systems

Methods, systems and computer program products provide assignment of ports for connections originated by multiple application instances executing on different data processing systems utilizing a common network address. An indication of available ports for the common network address is provided to each of the different data processing systems executing the multiple application instances. A port identified as available is selected as a port for a connection utilizing the common network address.

Methods, systems and computer program products for port assignments of multiple application instances using the same source IP address

Methods, systems and computer program products provide for transferring a virtual Internet Protocol address (VIPA) between communication protocol stacks including at least a first communication protocol stack and a second communication protocol stack. Ownership of the VIPA is established at the second communication protocol stack and connections to the first communication protocol stack utilizing the VIPA are maintained. TCP/IP messages for the connections to the first communication protocol stack are routed through the second communication protocol stack so as to non-disruptively move the VIPA from the first communication protocol stack to the second communication protocol stack.

Methods, systems and computer program products for non-disruptively transferring a virtual internet protocol address between communication protocol stacks

Cluster-wide port assignments in a workload distribution system in a cluster of data processing systems are provided such that a unique port may be provided for a socket. Establishing ports associated with workload distribution is also provided by allowing either enumerated ports and/or all ports of a common IP address to be associated with workload distribution. Workload distribution is provided in a cluster of data processing systems where a single IP address may be associated with a plurality of data processing systems and/or protocol stacks in the cluster of data processing systems and a port of the single IP address may be shared among multiple data processing systems and/or protocol stacks or may be unique among the plurality of data processing systems and/or protocol stacks.

Systems, method and computer program products for cluster workload distribution without preconfigured port identification by utilizing a port of multiple ports associated with a single IP address

Methods, systems and computer program products provide for recovery from a failure of a primary routing communication protocol stack which routes communications over connections to a virtual Internet Protocol address (VIPA) and at least one port associated with the VIPA to a plurality of other communication protocol stacks associated by the primary routing stack and with the VIPA and the at least one port. At least one backup routing communication protocol stack is identified. The backup routing stack stores an identification of stacks associated with the VIPA and the at least one port as candidate target stacks. Messages are received at the backup routing stack identifying stacks having application instances bound to the VIPA and listening on the at least one port as current actual target stacks. Connections to the VIPA and the at least one port for current actual target stacks are also identified and a routing table constructed from the received messages so as to provide routing path information to current actual target stacks associated with the connection. Messages on the connections to the VIPA and the at least one port are routed through the backup stack utilizing the constructed routing table.

Methods, systems and computer program products for failure recovery for routed virtual internet protocol addresses

Method and apparatus for sending data from one partition to a second partition within a logically partitioned computer. In a data processing system having multiple logical partitions, a send queue is established in the first logical partition, and a receive queue is established in the second logical partition. The send queue is registered in the send queue in a lookup table available to all of the logical partitions. The send queue is registered using as a key the logical partition identification of the first logical partition and the subchannel number (LPAR-ID.SUBCHANNEL#) of the subchannel assigned to the partition. The receive queue is registered in the lookup table using as a key, the internet protocol address of the receive queue in the second partition. A send instruction from the first logical partition is executed which interrogates the lookup table using the LPAR-ID.SUBCHANNEL# key to locate the send queue and IP address key to locate the receive queue, and sends the data in the send queue in the first logical partition to the receive queue in the second logical partition. This method and apparatus provides that discrete servers may be used in each logical partition, and data may be transferred between while maintaining security between the logical partitions.

Communications between partitions within a logically partitioned computer

Methods, systems and computer program products are provided for transferring a Virtual IP Address (VIPA) from a first application instance to a second application instance executing on a cluster of data processing systems having a plurality of communication protocol stacks. A list of dynamic VIPAs is distributed among the protocol stacks and a hierarchy of backup communication protocol stacks for the dynamic VIPAs is determined based on the list of dynamic VIPAs. Upon receiving notification of failure of the first stack the second stack evaluates the hierarchy of backup stacks to determine if it is the next stack in the hierarchy for the VIPA associated with the first application instance. If so, then the VIPA associated with the first application instance is transferred to the second communication protocol stack associated with the second application instance.

Methods, systems and computer program products for automated movement of IP addresses within a cluster

Methods, systems and computer program products are provided which provide connection information about connections to communication protocol stacks in a cluster of data processing systems by distributing location information between communication protocol stacks in a group of stacks within the cluster. The distributed location information about addresses and operating system images or data processing systems of communication protocol stacks is utilized to generate a list of addresses corresponding to stacks in the group so that this information may be provided to applications requesting the information from the communication protocol stacks.

Methods, systems and computer program products for providing network connection information in a cluster of data processing systems

In a word processing system, a text stream is stored on a direct access storage device (DASD) for recall and editing. The text stream is organized into a document that is stored on the DASD as a data set. The data set comprises an index portion and a portion containing text records. The index portion is divided into nodes, of which the primary node is called the root node. A method called shadowing is introduced to protect index nodes from the loss of significant data. Status indicators are defined and maintained so that access to a data set for normal system operation may be prevented where there is significant chance that the data set index has errors or is inconsistent to the extent that the normal high performance access method routines may not be able to use the index or may cause inadvertent loss of valid text data. Recovery mechanisms enable all of the data on the DASD which can be located by indexes to be restored after normal terminations during complex data set index updates or when intermediate nodes in large data sets are lost due to storage media input physical errors.

Disk volume data storage and recovery method

A word processing system is disclosed in which a text stream is stored on a direct access storage device (DASD) for recall and editing. The text stream is organized into a document, which is stored on the DASD as a data set. The data set consists of an index portion and a portion containing text records. The index portion is divided into nodes, of which the primary node is called the root node. When the system operator designates a data set for use, the root node is fetched and retained in a system memory to eliminate the need to fetch it from the DASD for every access. When sequential records are accessed, such as during a document duplication operation, the index is fully searched only for the first of the sequential records. Subsequent records are located through index nodes already in the system memory, with subsequent index nodes being fetched from the DASD only when the last index entry in the current index node has been processed. When sequential records are written to the DASD, the index is fully searched only for the insertion point for the first record. For subsequent records, the same index node is updated until there is insufficient room for index entry for the latest record written to the DASD. At that time the index node is split and index entries at higher levels of the index are updated in a similar manner.

Method for accessing a data set in a word processing system

In a word processing system storing a text stream on a direct access storage device (DASD) for recall and editing, the text stream is organized into a document, which is stored on the DASD as a data set. The data set consists of an index portion and a portion containing text records. The index portion is divided into nodes, of which the primary node is called the root node. A volume which refers to the full contents of the DASD can be formatted to contain multiple zones, each of which is logically independent of all other zones, when the contents of a storage volume can be divided conceptually into portions the uses of which are logically independent. Each zone has its own media allocation map, and data sets whose root is allocated within a given zone have all allocations within that same zone. If more than one data set with a given data set type exists on the storage volume, all of those data sets and the data set director for that data set type are allocated within a single zone. For each zone, a recovery required indicator is provided so that zones not requiring recovery need not be scanned when volume recovery is being performed. When only one zone is being recovered, the logical sector number of the root node referenced by the index element in the anchor is used to determine whether or not the data set lies in the zone being recovered, and to skip data sets that do not lie in the zone.

Zone partitioning in volume recovery system

The specification discloses a method for adjusting text in a word processing system when the insert mode is completed. An indicator is set to indicate a change active condition if a change is made to the current line of text when an establish insert mode key is depressed by the operator. The indicator is interrogated to determine if a change active condition has been indicated when a terminate insert mode key is later depressed by the operator. When a change active condition is detected, the line end codes for the current line of text are adjusted to provide an adjusted line. The adjusted line is then interrogated to determine whether a paragraph boundary or page end code is included. If no paragraph boundary or page end code is included, the adjusted line is interrogated to determine whether a change was made in the line end code. Each succeeding line following the adjusted line is adjusted until a paragraph boundary or page end code is encountered or no change in the line end code is made for that line.

Growing community of inventors

Raleigh, NC, United States of America

John Andrew Aiken, Jr