Oracle International Corporation

Rajiv Wickremesinghe

Prasad V Bagal

Harish Nandyala

Richard L Long

Shie-rei Huang

Sameer Arun Joshi

David Friedman

Lavina Jain

Darshan Nagarajappa

William Havinden Bridge, Jr

Hanlin Daniel Chien

Dungara Ram Choudhary

Mahanteshwar Chimangala Maheshwar

Hemanth Pikkili

Sahaj Agarwal

Willliam H Bridge, Jr

Embodiments enable a database management system (DBMS) to manage two levels of disk failure groups. These two levels of redundancy are achieved by grouping the disks of the disk group for the DBMS into two levels of failure groups (i.e., 'data sites' each containing two or more 'failure groups' of disks). This system of disk grouping allows a DBMS to potentially tolerate the loss of both an entire first site and part of a second site. Such a DBMS uses a multi-level voting system, based on both failure group-level votes and site-level votes, to identify the current version of administrative data structures (ADS) that store key administrative data. In addition to data sites that store database data, the DBMS includes a quorum site with a single quorum failure group that stores a copy of the ADS. The quorum site contributes a site-level vote during a multi-level voting event.

Hierarchical fault tolerance in system storage

A method and apparatus for booting a clustered management system is provided where the configuration information required to boot the system is stored within the system itself. The techniques cope with several issues including finding the configuration data on the physical disk before the logical disk system is established, coping with replicated copies of configuration data, network partitions and disk drive failures that may prevent a booting node to have perfect knowledge of the state of the entire system, and active nodes modifying or moving configuration data while the booting node is searching for a consistent, correct copy of configuration data.

Bootstrapping server using configuration file stored in server-managed storage

A method, system, and computer program for high-availability database systems. The method commences by replicating (e.g., from a first server to a second server) a data structure having entries describing resource locks and client states. The replication can be performed to any number of servers. Upon detecting a failure of a connection between the first server and a client to which the first server is connected, the system establishes a new connection between the second server and the client by updating the replicated data structure to describe new connection between the second server and the client and updating the entries of the replicated data structure to reflect the new connection and the then current aspects of the resource locks and client states. The client can be a database instance, and the act of replicating can be accomplished using a broadcast channel. The servers can be organized in a computing cluster.

Systems and methods to retain and reclaim resource locks and client states after server failures

A method, system, and computer program product for restoring blocks of data stored at a corrupted data site using two or more mirror sites. The method commences by receiving a trigger event from a component within an application server environment where the trigger event indicates detection of a corrupted data site. The trigger is classified into at least one of a plurality of trigger event types, which trigger event type signals further processing for retrieving from at least two mirror sites, a first stored data block and a second stored data block corresponding to the same logical block identifier from the first mirror site. The retrieved blocks are compared to determine a match value, and when the match value is greater than a confidence threshold, then writing good data to the corrupted data site before performing consistency checks on blocks in physical or logical proximity to the corrupted data site.

Detection of logical corruption in persistent storage and automatic recovery therefrom

Techniques are provided for assigning read requests to storage devices in a manner that reduces the likelihood that any storage device will become overloaded or underutilized. Specifically, a read-request handler assigns read requests that are directed to each particular item among the storage devices that have copies of the item based on how busy each of those storage devices is. Consequently, even though certain storage devices may have copies of the same item, there may be times during which one storage device is assigned a disproportionate number of the reads of the item because the other storage device is busy with read requests for other items, and there may be other times during which other storage device is assigned a disproportionate number of the reads of the item because the one storage device is busy with read request for other items. Various techniques for estimating the busyness of storage devices are provided, including fraction-based estimates, interval-based estimates, and the response-time-based estimates. Techniques for smoothing those estimates, and for handicapping devices, are also provided.

Assigning read requests based on busyness of devices

Growing community of inventors

Belmont, CA, United States of America

Rajiv Wickremesinghe