Compaq Computer Corporation, Inc.

Digital Equipment Corporation

Hewlett-packard Development Company, L.p.

Compaq Information Technologies Group, L.p.

Johannes Klein

Sitaram V Lanka

Albert C Gondi

Robbert C Van Der Linden

Raj K Rathee

Roger J Hansen

William J Carley

Hansjorg Zeller

Keith B Evans

Sameer Arun Joshi

Anoop Sharma

Edward C Cheng

Alberto Lutgardo

Dora L Lee

Ramprasad K L Sripada

Edward Y Chang

Francis R Upton, Iv

John S De Roo

Edward S Lu

Methods and systems that provide a resource manager the ability to register with a transaction manager that it needs to be &ldquo;pre-prepared&rdquo; for a transaction. The transaction manager will issue a pre-prepare notification to any resource manager so registered, and will delay phase one of the two-phase commitment (2PC) protocol until all pre-prepare-registered resource managers have responded that they are ready for commitment processing.

Volatile resource manager with pre-prepare notification

A transaction processing system manages an information collection by employing a two-phase transaction commit protocol. Transaction participants are provided a unique identifier for use in communicating with the transaction management facility in a distributed environment. If the participant is lost, a backup participant is created, and provided with the same identifier as the failed participant to carry on with the transaction in place of the failed participant.

Method and apparatus for handling failures of resource managers in a clustered environment

The SQL compiler and SQL executor in a database management system are extended to process queries requiring streaming mode processing of data stored in a table. A scan operator performs table access functions during execution of the streaming mode query. The scan operator first performs an initial scan to access rows in a specified database table, and then performs a delta scan to access new rows added to the table, as well as rows modified by other queries. The scan operator continues to process new data added to the table until the initiating user or application closes the associated cursor. A set of data structures are provided to keep track of active scan operators, including a session control block that includes fields for keeping track of whether the scan operator is performing an initial scan or a delta scan. The session control block also includes, for streaming mode scan operators, a delta scan list of new and modified rows to be processed in the delta scan. When a transaction terminates, any streaming mode scan operators whose operation was suspended because of rows locked by the transaction are automatically rescheduled for execution.

System and method for performing database operations on a continuous stream of tuples

Incorporation of a collateral process as a participant in a transaction is made possible by a method and system in accordance with the present invention. Typically, after the transaction is initiated, the collateral process is called and then is registered as a participant,in the transaction. A prepare signal is sent to each registered collateral process when end stage of the transaction is reached. Then, a ready signal is received from the collateral process if the collateral process is completed successfully; and an abort signal is received from the collateral process if the collateral process does not complete successfully or a violation is detected. If a ready signal is received, a commit record is written to a log, and a commit signal is sent to each registered collateral process. In response to the commit signal, a forgotten signal is received from each registered collateral process.

Incorporating collateral and concurrent activity in a data processing transaction

A protocol for a transaction involving two homogeneous or two heterogeneous computing systems involves starting a transaction on one of the two systems, sending a request for participation in the transaction to an application resident on the other of the two systems, together with an identification and address of the transaction. Upon receipt of the request, the application will initiate a subordinate transaction through a resident (subordinate) transaction manager. The subordinate transaction manager will notify the Beginner transaction manager and at the same time cause the application to start work on the request. Later, the subordinate transaction, through the subordinate transaction manager participates in a two-phase commit protocol that concludes the transaction to ensure that all changes effected by the transaction are done, or none are done, i.e., the transaction is aborted.

Method and apparatus for conducting a transaction between homogeneous and/or heterogeneous transaction processing systems using asynchronous pull of a transaction transfer

Systems and Methods for identifying in the gateway processes or application processes whether an imported transaction is a loop-back transaction, and if so performing work for the imported transaction under the parent transaction. Upon receiving a work request from a remote transaction processing system, an endpoint identifier included as part of the work request is analyzed to determine whether the work request is associated with a parent transaction initiated at the local transaction processing system. If the endpoint identifier indicates that the work request is associated with a transaction initiated in the local transaction processing system, a global transaction identifier included as part of the work request is analyzed to determine the identity of the parent transaction. If the work request includes a remote TIP URL, the application process that receives the request is configured to analyze the global transaction identifier and determine the identity of the parent transaction. If the remote transaction processing system sends a push request, the gateway process that receives the request is configured to analyze the global transaction identifier and determine the identity of the parent transaction. Work is thereafter performed according to the work request under the parent transaction such that a new transaction is not initiated.

Systems and methods for the detection of a loop-back of a transaction

Systems and methods for supporting and maintaining a distributed global map of transaction identifiers at the gateway processes using a hashing algorithm configured on each application process to access the global maps. A global map of transaction identifiers that associates global transaction identifiers with remote local transaction identifiers is maintained at each gateway process. When an application process performing work for a particular transaction desires to export the transaction to a remote node, a hashing function configured on the application process is applied to the global transaction identifier associated with the particular transaction. Application of the hashing function to the global transaction identifier identifies one of the gateway processes. The global transaction identifier is stored to the global map associated with that gateway process. When the remote transaction manager associated with the remote node responds with a local transaction identifier for a local transaction initiated at the remote node, the local transaction identifier is stored to the identified gateway's global map in association with the global transaction identifier. Each application process is configured with the same hashing function so that the same gateway process will always be identified given a particular global transaction identifier. If the same or another application process desires to export to a remote node a transaction that has already been exported, the hashing function on that application process identifies the same gateway through which the transaction was originally exported.

Process of maintaining a distributed map of transaction identifiers and using hashing to access these maps

The SQL compiler and SQL executor in a database management system are extended to execute statements for dequeuing rows from a database table and statements that both updating tuples and return those tuples to a calling application. During execution of a select statement that includes an embedded update or delete operation, a table access operator accesses a defined range of rows in a database table. The table access operator receives from a calling application a request for rows dequeued from the database table. The table access operator responds to the requests for rows by returning to the calling application qualifying rows, if any, from the database table and also by performing the embedded delete or update operation. The SQL executor returns control to the calling application only after all rows deleted or updated by the table access operator have been sent to the calling application.

Database management system and method for dequeuing rows published to a database table

A database management system includes an execution engine that, upon rollback of an aborted transaction, has the ability to set fields of the rows that are updated or deleted by the transaction prior to aborting. In particular, when a select, update or delete statement includes a &ldquo;set on rollback&rdquo; clause, the respective table access operator of the execution engine modifies the &ldquo;before image&rdquo; of each row that is stored by the transaction log manager. The modified before image includes the field updates specified by the set on rollback clause. If a transaction associated with the statement aborts, when the transaction log manager restores the before images of each row that was deleted or updated by the transaction, the restored before images include the field updates specified by the set on rollback clause.

Database management system and method for updating specified tuple fields upon transaction rollback

In a database management system, when the table to be accessed by a statement is partitioned, a respective partition scan operator is used to access rows of each respective database table partition. A fan out operator receives requests from a calling application to access rows from table, and responds by sending requests for rows to the partition scan operators. It receives rows of the table from the partition scan operators and sends the received rows of the database table to another node in the execution tree. Each of the partition scan operators responds to the requests for rows by returning to the fan out operator qualifying rows, if any, from the corresponding table partition. If the statement being executed includes an embedded delete or update operation, the qualifying rows are deleted or updated at the same time that the qualifying rows are returned to the fan out operator. Control is returned to the calling application only after all rows deleted or updated by the partition scan operators have been sent to the calling application. When the statement being executed is a streaming mode query, the fan out operator sends non-blocking requests to the partition scan operators so that when one access to one partition is blocked, rows from another partition are processed.

Database management system and method for accessing rows in a partitioned table

A database management system is extended to process SQL statements so as to skip over rows that are locked in conflicting modes. Normally, a table access operator is blocked when it attempts to access a row that is locked in a conflicting mode, such as a write lock mode, that is inconsistent with the read or write access required by the operator. Generally, the lock in the conflicting mode will be held by a transaction other than the transaction associated with the operator in question. When the SQL statement being executed uses the &ldquo;skip conflict&rdquo; syntax provided by the present invention, the operator skips over rows that are locked in conflicting modes, which would otherwise cause the operator to suspend operation. Furthermore, if the operator is operating in streaming mode, a key that identifies such skipped rows is added to the operator's list of rows to be processed at a later time.

System and method for performing database operations and for skipping over tuples locked in an incompatible mode

Systems and methods for ensuring that two different subordinate transactions will not be created at any given transaction processing node for the same parent transaction. Upon receiving a work request from a remote transaction processing system, the local transaction manager creates a local subordinate transaction associated with the parent transaction. A local transaction identifier that identifies the local subordinate transaction is also created. The local identifier includes an indicator, such as a one-bit flag, which identifies the local subordinate transaction as an imported transaction. The global identifier for the parent transaction is stored to a map in association with the transaction identifier of the new local subordinate transaction. If a work request associated with an imported transaction is sent to a remote transaction processing system, the global transaction identifier of the parent transaction is retrieved from the map and included in the work request so that the receiving system will know that the work request is associated with the parent transaction. In this manner, the receiving system will know not to start a new subordinate transaction on behalf of the parent transaction if a transaction subordinate to the parent transaction has already been started on that system.

Detection of an imported transaction for finding the global transaction identifier

A transaction processing system, including multiple processor units communicatively interconnected, manages information collection by employing a distributed transaction management facility to track and make consistent changes. When each transaction is started, a data structure is created that maintains information concerning the transaction. Included in the data structure is the identity of all processor units having resources involved in the transaction. Should a processor unit fail, and the transaction management facility is notified of that failure, the data structures of all pending transactions will be examined to see if the failed processor unit had a resource that was a participant in the corresponding transaction. If so, the transaction management facility can then make a decision as to whether or not to abort the transaction.

Growing community of inventors

San Francisco, CA, United States of America

Johannes Klein