Oracle International Corporation

David Brower

Roger J Bamford

Sashikanth Chandrasekaran

William Bridge

Neil MacNaughton

Wilson Wai Shun Chan

Vinay H Srihari

Carol L Colrain

Jonathan Creighton

Angelo Pruscino

William Havinden Bridge, Jr

Tak Fung Wang

Andrey Gusev

Juan R Loaiza

Meichun Hsu

Tirthankar Lahiri

Garret Frederick Swart

Boris Klots

James William Stamos

Rabin Sugumar

Lakshminarayanan Luxi Chidambaran

Christopher Angel Kantarjiev

Debashish Chatterjee

Ajit Mylavarapu

Wilson Wai Chan

Benny Souder

Neil J S MacNaughton

Umesh Panchaksharaiah

Sarat B Kakarla

Raj K Kammend

Daniel Semler

John Leys

Gary D Young

Yuri Sharonin

William Jr Bridge

Neil Macnaughton

Neil J S Macnaughton

An elector is provided using shared storage to allow a cluster of participant nodes to gracefully convert failed voters to non-voters until there are only two voters. The cluster can then tolerate the failure of either of those two remaining participant nodes. A tiny amount of shared storage is used as a voter without an object store. This results in a cluster with only one running node that is an exclusive leader. An exclusive leader is then the only participant node that is a voter, and the exclusive leader can commit log records without replicating them to any other members of the cluster. When there is an exclusive leader, there is only one remaining node with a current copy of the object store.

Tolerating server failures in a consensus protocol down to one surviving server using shared storage as a voter

A shared-nothing database system is provided in which parallelism and workload balancing are increased by assigning the rows of each table to 'slices', and storing multiple copies ('duplicas') of each slice across the persistent storage of multiple nodes of the shared-nothing database system. When the data for a table is distributed among the nodes of a shared-nothing system in this manner, requests to read data from a particular row of the table may be handled by any node that stores a duplica of the slice to which the row is assigned. For each slice, a single duplica of the slice is designated as the &#x201c;primary duplica&#x201d;. All DML operations (e.g. inserts, deletes, updates, etc.) that target a particular row of the table are performed by the node that has the primary duplica of the slice to which the particular row is assigned. The changes made by the DML operations are then propagated from the primary duplica to the other duplicas (&#x201c;secondary duplicas&#x201d;) of the same slice.

System and method for rapid fault detection and repair in a shared nothing distributed database

Techniques are provided for managing caches in a system with multiple caches that may contain different copies of the same data item. Specifically, techniques are provided for coordinating the write-to-disk operations performed on such data items to ensure that older versions of the data item are not written over newer versions, and to reduce the amount of processing required to recover after a failure. Various approaches are provided in which a master is used to coordinate with the multiple caches to cause a data item to be written to persistent storage. Techniques are also provided for transferring data items and locks associated with the data items from one node to another.

Methods to perform disk writes in a distributed shared disk system needing consistency across failures

A method for locking resources, including: receiving, by an accelerator, a first request from a first client to lock a first resource; evaluating, by a computer processor of a server, a hash function using an identifier of the first resource as an input to the hash function; identifying, by the computer processor and based on evaluating the hash function, a first hash bucket in a shared memory residing in a physical memory of the server; detecting that the first hash bucket is occupied; and sending the first request to a master lock monitor residing in a user space of the server based at least on detecting that the first hash bucket is occupied.

Lock acceleration

A method and computer-readable storage representing resources in a cluster by a plurality of attribute-value pairs that together are part of a 'resource profile,' in which each attribute-value pair defines all, or a portion of, a management policy that applies to the resource. A clusterware manager configures a resource according to the resource profile, and follows a resource profile syntax that specifies a runtime value for the resource in which an actual value is substituted at runtime.

Special values in oracle clusterware resource profiles

A cluster management system dynamically assigns a category to a server node based on the server node's attributes. The system can detect a change in value of a server attribute of a first server node. Then, in response to detecting the change in value of the server attribute, the system identifies a plurality of server attributes for the first server node, and dynamically assigns to the first server node a category that matches the server attributes. Further, the system can add the server node to a server pool that has a matching category with the server node.

Assigning server categories to server nodes in a heterogeneous cluster

A method for offloading computation flexibly to a communication adapter includes receiving a message that includes a procedure image identifier associated with a procedure image of a host application, determining a procedure image and a communication adapter processor using the procedure image identifier, and forwarding the first message to the communication adapter processor configured to execute the procedure image. The method further includes executing, on the communication adapter processor independent of a host processor, the procedure image in communication adapter memory by acquiring a host memory latch for a memory block in host memory, reading the memory block in the host memory after acquiring the host memory latch, manipulating, by executing the procedure image, the memory block in the communication adapter memory to obtain a modified memory block, committing the modified memory block to the host memory, and releasing the host memory latch.

Method and system for offloading computation flexibly to a communication adapter

Techniques are provided for managing caches in a system with multiple caches that may contain different copies of the same data item. Specifically, techniques are provided for coordinating the write-to-disk operations performed on such data items to ensure that older versions of the data item are not written over newer versions, and to reduce the amount of processing required to recover after a failure. Various approaches are provided in which a master is used to coordinate with the multiple caches to cause a data item to be written to persistent storage. Techniques are also provided for managing checkpoints associated with the caches, where the checkpoints are used to determine the position at which to begin processing recovery logs in the event of a failure.

Techniques are provided for managing access to resources that are shared by a cluster of nodes. One aspect of the invention is to require verification that a quorum is a valid quorum prior to enforcing the access rights specified by a node that is a member of the group of nodes holding the quorum. Further, the verification that a quorum is a valid quorum prior to enforcing access rights occurs only if the quorum associated with the specified access rights is suspected to be invalid.

Method for fencing shared resources from cluster nodes

Managing checkpoint queues in a multiple node system

A composite resource is established that includes a plurality of members. Each of the members is capable of providing a comparable service. A coordinator monitors a state of each member of the composite resource. A component requests the service from the coordinator. The coordinator arranges for the service to be provided to the component by a particular member of the composite resource. When the particular member ceases to be active, the service is automatically provided to the component by another member in the composite resource. A state of the composite resource is maintained independently of the state of each member in the composite resource.

Managing a high availability framework by enabling and disabling individual nodes

Approaches are used for efficiently and effectively managing the dynamic allocation of resources of multi-node database systems between services provided by the multi-node database server. A service is a category of work that is hosted on the database server. The approaches manage allocation of resources at different levels. For services that use a particular database, the performance realized by the services is monitored. Resources assigned to the database are allocated between these services to ensure performance goals for each are met. Resources assigned to a cluster of nodes are allocated between the databases to ensure that performance goals for all the services that use the databases are met. Resources assigned to a farm of clusters are assigned amongst clusters based on service level agreements and back-end policies. The approach uses a hierarchy of directors to manage resources at the different levels.

Hierarchical management of the dynamic allocation of resources in a multi-node system

Real composite objects for providing high availability of resources on networked systems

Events of interest are detected in order to manage a high availability framework. In a framework in which a plurality of components are executing, the components are periodically polled to detect occurrence of the event of interest. A monitor is also established for one or more of the components. After the first component causes the event of interest to occur, the monitor communicating the event of interest to the framework without waiting for the framework to poll the first component.

Detecting events of interest for managing components on a high availability framework

Growing community of inventors

Alamo, CA, United States of America

David Brower