Patent Grant
9092460
A portion of the disclosure of this patent document contains material which is subject to copyright protection. The copyright owner has no objection to the facsimile reproduction by anyone of the patent document or the patent disclosure, as it appears in the Patent and Trademark Office patent file or records, but otherwise reserves all copyright rights whatsoever.
The present invention is generally related to computer systems and software such as middleware, and is particularly related to a system for managing connections to a clustered database.
In the context of an application server, such as the Oracle Weblogic Server, users can configure database connectivity in the application server by configuring data sources. An application on the application server can look up a particular data source using a directory service, such as the standard Java Naming and Directory Interface (JNDI), and then request a database connection. When finished with the connection, the application can disconnect the database connection through the application server. Both application server administrators and software developers/programmers can create data sources. These are the generally areas that embodiments of the invention are intended to address.
In accordance with an embodiment, a system and method is provided to support using a data source to connect an application server with a clustered database. The clustered database includes a plurality of database instances and is associated with a notification service. The notification service can be used by the clustered database to broadcast notifications that describe a state change in the plurality of database instances. The data source includes a connection pool, which manages a set of connections to the plurality of database instances in the clustered database. The data source operates to register with the notification service to receive notifications regarding the change of the clustered database, wherein the application server operates to configure and manage connections to the clustered database, adaptively according to any state changes of the clustered database at run time.
The present invention is illustrated, by way of example and not by way of limitation, in the figures of the accompanying drawings in which like references indicate similar elements. It should be noted that references to “an” or “one” or “some” embodiment(s) in this disclosure are not necessarily to the same embodiment, and such references mean at least one.
Additionally, the description of various embodiments of the invention provided herein use Oracle Real Application Clusters (RAC) database system as an illustrative example of a clustered database environment. It will be apparent to those skilled in the art that other types of clustered database environments can be used without limitation.
A clustered database or a database cluster can comprise multiple interconnected computers or servers that appear as if they are one server to end users and applications. Unlike a single-instance database, which has a one-to-one relationship between the database and the instance, a clustered database has a one-to-many relationship between the database and instances. For example, an Oracle RAC database system enables a user to cluster Oracle databases, using Oracle Clusterware infrastructure to bind multiple servers together, so that they operate as a single system. The Oracle RAC database system can have many instances, all of which access one database. The combined processing power of the multiple servers can provide greater throughput and scalability than what is available from a single server.
Gridlink Data Source
In accordance with an embodiment, a system and method is provided to support using a data source, referenced to herein as a “Gridlink data source”, to connect an application server with a clustered database. The clustered database includes a plurality of database instances and is associated with a notification service. The notification service can be used by the clustered database to broadcast notifications that describe a state change in the plurality of database instances. The data source includes a connection pool, which manages a set of connections to the plurality of database instances in the clustered database. The data source operates to register with the notification service to receive notifications regarding the change of the clustered database, wherein the application server operates to configure and manage connections to the clustered database, adaptively according to any state changes of the clustered database at run time.
In accordance with an embodiment, the gridlink data source can use a single data source configuration that represents a service targeted to a database cluster. The gridlink data source can respond to notification events to provide fast connection failover, runtime connection load balancing and database instance graceful shutdown. Additionally, distributed transaction (XA) affinity can be supported at the global transaction Id level.
For example, the Oracle RAC database system can support client notification to disseminate information about the state of the database cluster. A JDBC data source configuration allows for the specification of service URLs to provide RAC connectivity. The JDBC data source configuration also includes monitoring support that provides statistics and manageability of JDBC connections to the RAC cluster.
In accordance with an embodiment, the Gridlink data source addresses the shortcomings of the multi data source solution, by leveraging the capabilities of database cluster notifications to provide better overall connectivity in the form of simpler configuration, faster response to a database node failure, better utilization of database cluster resources, and improved runtime monitoring and management. In addition, mixed configurations of the Gridlink data sources and multi data sources can be supported.
In accordance with an embodiment, the management of the connections in the connection pool is based on static settings 214 configured on the connection pool, such as min/max capacity, timeouts, etc., and real time information about the connection pool in event messages 210 received from the notification service 206 that advises the data source of any state changes within the database cluster. When an application 207, 208, or 209 requests a connection from the data source, a suitable connection is selected from the connection pool and supplied to the application based on the load balancing information the connection pool has received and the current distributions of connections in use from the pool.
The application server can register with the notification service to receive notifications, such as database event messages, and therefore quickly become aware of any state changes in a clustered database. Using these state change notification events, the application server can intelligently adapt its connection pools so that the system can provide continuous, reliable and efficient access to the clustered database.
In accordance with an embodiment, connection polling can be applied as an alternative to the notification service, when the notification service is not configured or it is not operating correctly. When connection polling is used, the system can determine the viability of individual JDBC connections and detect changes in the clustered cluster topology, by performing SQL operations on individual connections.
Comparing with the notification service approach, the connection polling approach comes at the expense of additional runtime overhead, and potentially delayed detection of database instance node failures. Also, the connection polling approach potentially suffers false positives that can result in the unnecessary disablement of the data source pool and termination of valid connections that may be in use by applications.
Also as shown in
For example, Oracle Database services can be logical abstractions for managing workloads in Oracle Database. The database services can provide a single system image for workloads, prioritization for workloads, performance measures for real transactions, and alerts and actions when performance goals are violated. The database services can enable database administrators to configure a workload, administer workloads, enable/disable workloads, and measure workloads as a single entity. The database services can further divide workloads into logically disjoint groupings. Each service represents a workload with common attributes, service-level thresholds, and priorities.
In accordance with an embodiment, the configuration of a Gridlink data source can leverage JDBC descriptor beans that are persisted. The JDBC descriptor beans can identify a Gridlink data source and specify database notification service client configuration information. The JDBC descriptor beans can use an XML file to enable Gridlink data source functionalities. The following is an exemplary XML file for configuring a Gridlink data source that is connected to an Oracle RAC database system.
Load Balancing Advisory Event
In accordance with an embodiment, the clustered database can provide a runtime load balancing service to distribute connections across the database instance based on performance goals set by a database administrator (DBA), in order to provide better throughput and more efficient use of resources. The load balancing advisory service issues events that advise clients on the current state of the cluster including advice on where to direct connections.
The use of the runtime load balancing advisory service 506 in conjunction with the single data source 503 can simplify initial configuration and ongoing maintenance of the data source. Using the Gridlink data source system, database topology information is not needed for the application server to support load balancing, since the database directs the load balancing activities. The Gridlink data source system helps the initial construction of a data source by reducing the amount of information required to correctly configure the middle tier. Additionally, as the database tier may change over time and services may be relocated onto new or different database instances, the Gridlink data source system requires no configuration change to reflect the changed database topology.
For example, as shown in
In accordance with an embodiment, in addition to readjusting connections in the connection pool based on the load-balancing information, existing connections can be given out to application components according to the load-balancing percentages/weights when there are idle connections available.
Database State Change Event
In accordance with an embodiment, the Gridlink data source system allows the application server to adaptively respond to state changes in the database cluster, such as handling outages by immediately retracting, closing and discarding connections to database instances that have been stopped or taken out by an unplanned outage. As shown in
As shown in the above example, using database state change events, the Gridlink data source system does not need to periodically poll the connections to ensure they are valid, or affecting uninvolved connections to surviving nodes. This can lessen the reliance on testing of connections to ensure that applications are not given invalid connections and the application server are given the information to quickly free applications from connections that become invalid because of database node failures without unnecessary delays.
In addition to handling unexpected database outage, the Gridlink data source allows an application server to proactively reapportion its set of connections to support scenarios where new database instances are added or are restarted after an outage. This allows the application server to make full use of the resources within the clustered database. Furthermore, using the database service model, the system allows database administrators to make changes to the database service/instance allocations, which are then seamlessly applied through the affected application server connection pools without a need to make configuration changes to the connection pool configuration. It also removes the need to create complex arrangements of multiple data sources to represent a dedicated instance of the clustered database.
Furthermore, using the database state change events, the Gridlink data source can provide fast connection failover capabilities and responds to clustered database service and node events, such as {UP, DOWN} events, to ensure that the reserve of physical connections in the pool are always pointing to a valid database node. Furthermore, the Gridlink data source ensures that the reserve of physical connections is well distributed across the available database nodes. The Fast Connection Failover behavior can be enabled as a configuration setting on the data source. Metrics can be made available to allow administrators to monitor and review what action an application server has taken on its data sources upon receipt of database notification events.
Distributed Transaction (XA) Affinity
In accordance with an embodiment, distributed transaction (XA) affinity is used to ensure that all database operations performed on a database cluster within the context of a global transaction are directed to the same database instance. Affinity can be established based on the global transaction id, instead of by individual data source, to ensure that connections obtained from different data sources that are configured for the same database cluster are all associated with the same database instance. The two-phase commit optimization can be supported by the Gridlink data source and can also participate in XA affinity.
Database Graceful Planned Shutdown
In accordance with an embodiment, a graceful planned shutdown occurs when a database node/service is targeted for a shutdown operation, with the corresponding issuance of a database event indicating the shutdown has been requested.
In order to support graceful planned shutdown of the database, the application server may not immediately abort connections 821 that are in use when it detects that the database shutdown target is no longer accepting new connections. Instead, the Gridlink data source 803 allows any in progress transactions to complete before closing and recreating the physical connections, while cleaning up idle connections so that new requests for connections are not sent the database target in active shutdown mode.
The data source can detect the active connections that are connected to the database shutdown target, and mark them so that they are closed and recreated when the associated transaction is complete and the connection is returned to the connection pool 805. An exception may not be created immediately on the receipt of the shutdown event or detection of disallowed connection request, in order to allow any in progress transactions to complete. Additionally, any idle connections in the pool that are connected to the database shutdown target can be preemptively closed. The system also allows the data source to route new requests around the database shutdown target, while allowing in progress transactions to complete to fulfillment, enabling the database shutdown operation to be transparent to running applications.
Universal Connection Pool (UCP) Library
In accordance with an embodiment, a database connection module, such as a Universal Connection Pool (UCP) library, can be used to connect an application server to a clustered database. The UCP library is a client-side library from the perspective of the database server that is logically an extension of the JDBC driver. The database connection module can support high availability and performance capabilities for a clustered database, in addition to providing a generic JDBC connection pooling implementation.
In accordance with an embodiment, without a need to use the JDBC connection pooling feature provided by the UCP library, the Gridlink data source can connect to the database cluster and take advantage of other capabilities supported by the UCP library, such as fast connection failover, runtime connection load balancing and XA connection affinity.
In accordance with an embodiment, the connection module can utilize an application server timer and work managers for better utilization and management of resources such as Java Virtual machines (JVMs). In addition, an API can allow an application server to obtain clustered database advisories to more efficiently allocate connections across database instances and to better respond to changes in the database cluster topology. For example, the RAC capabilities of UCP can be refactored and exposed to the application server in the form of a RAC module API. This API can provide callback notifications in response to changes in the RAC cluster topology and advisories for runtime connection load balancing and connection affinity. The API also provides notification event callbacks that an application server can use for logging and other diagnostic purposes. The API can surface statistics counts and status information about RAC cluster nodes that can be used for RAC data source runtime monitoring.
In accordance with an embodiment, UCP can be designed as a standalone Java library that provides connection pooling and database integration features. As such, the UCP library can manage its own threads and timers. UCP can provide a timer and work manager SPI that allows the application server to plug in instances of application server timers and work managers for UCP's internal use when the UCP APIs are accessed within an application server process. This allows for increased control and runtime visibility of the threads and timers used by the UCP library. Additionally, the affinity capabilities provided by UCP can be leveraged to assign connections based on a global transaction Id even when different data sources are accessed on the same, and separate, application server instances.
Data Source Creation Wizard
In accordance with an embodiment, a data source creation wizard can support creating a Gridlink data source that is configured to connect with a clustered database, such as an Oracle RAC database. The creation of a Gridlink data source includes setting up database listener address information and notification service client configuration information that are needed for connecting to a database service.
For example, in order to connect with the Oracle RAC database, the data source creation wizard can provide two ways to specify the connection target. The first approach uses a connect_descriptor in JDBC URL format, which is shown as the following, for specifying different listeners for the data source.
The second approach is a wizard driven approach that allows each of the listener entries to be specified individually, from which the final JDBC URL can ultimately be constructed. The wizard driven approach enables a user to specify and test multiple listeners and notification service clients for creating and updating Gridlink data sources.
The present invention may be conveniently implemented using one or more conventional general purpose or specialized digital computer, computing device, machine, or microprocessor, including one or more processors, memory and/or computer readable storage media programmed according to the teachings of the present disclosure. Appropriate software coding can readily be prepared by skilled programmers based on the teachings of the present disclosure, as will be apparent to those skilled in the software art.
In some embodiments, the present invention includes a computer program product which is a storage medium or computer readable medium (media) having instructions stored thereon/in which can be used to program a computer to perform any of the processes of the present invention. The storage medium can include, but is not limited to, any type of disk including floppy disks, optical discs, DVD, CD-ROMs, microdrive, and magneto-optical disks, ROMs, RAMs, EPROMs, EEPROMs, DRAMs, VRAMs, flash memory devices, magnetic or optical cards, nanosystems (including molecular memory ICs), or any type of media or device suitable for storing instructions and/or data.
The foregoing description of the present invention has been provided for the purposes of illustration and description. It is not intended to be exhaustive or to limit the invention to the precise forms disclosed. Many modifications and variations will be apparent to the practitioner skilled in the art. The embodiments were chosen and described in order to best explain the principles of the invention and its practical application, thereby enabling others skilled in the art to understand the invention for various embodiments and with various modifications that are suited to the particular use contemplated. It is intended that the scope of the invention be defined by the following claims and their equivalence.
This application claims the benefit of priority on U.S. Provisional Patent Application No. 61/383,285, entitled “MIDDLEWARE MACHINE PLATFORM”, filed Sep. 15, 2010; and U.S. Provisional Patent Application No. 61/384,227, entitled “MIDDLEWARE MACHINE PLATFORM”, filed Sep. 17, 2010, each of which applications are herein incorporated by reference.
| Number | Name | Date | Kind |
|---|---|---|---|
| 5109384 | Tseung | Apr 1992 | A |
| 6424992 | Devarakonda et al. | Jul 2002 | B2 |
| 6895590 | Yadav | May 2005 | B2 |
| 6938085 | Belkin et al. | Aug 2005 | B1 |
| 7376953 | Togasaki | May 2008 | B2 |
| 7394288 | Agarwal | Jul 2008 | B1 |
| 7483374 | Nilakantan et al. | Jan 2009 | B2 |
| 7554993 | Modi et al. | Jun 2009 | B2 |
| 7991904 | Melnyk et al. | Aug 2011 | B2 |
| 8130776 | Sundararajan et al. | Mar 2012 | B1 |
| 8131860 | Wong et al. | Mar 2012 | B1 |
| 8260757 | Chatterjee et al. | Sep 2012 | B1 |
| 20020174136 | Cameron et al. | Nov 2002 | A1 |
| 20030014480 | Pullara et al. | Jan 2003 | A1 |
| 20030078958 | Pace et al. | Apr 2003 | A1 |
| 20030120822 | Langrind et al. | Jun 2003 | A1 |
| 20040177126 | Maine | Sep 2004 | A1 |
| 20040205771 | Sudarshan et al. | Oct 2004 | A1 |
| 20050021354 | Brendle et al. | Jan 2005 | A1 |
| 20050027901 | Simon et al. | Feb 2005 | A1 |
| 20050038801 | Colrain et al. | Feb 2005 | A1 |
| 20050094577 | Ashwood-Smith | May 2005 | A1 |
| 20050102412 | Hirsimaki | May 2005 | A1 |
| 20050223109 | Mamou et al. | Oct 2005 | A1 |
| 20050262183 | Colrain et al. | Nov 2005 | A1 |
| 20050262215 | Kirov et al. | Nov 2005 | A1 |
| 20060015600 | Piper | Jan 2006 | A1 |
| 20060031846 | Jacobs et al. | Feb 2006 | A1 |
| 20060129676 | Modi et al. | Jun 2006 | A1 |
| 20060143525 | Kilian | Jun 2006 | A1 |
| 20060176884 | Fair et al. | Aug 2006 | A1 |
| 20060209899 | Cucchi et al. | Sep 2006 | A1 |
| 20060248200 | Stanev | Nov 2006 | A1 |
| 20060294417 | Awasthi et al. | Dec 2006 | A1 |
| 20070198684 | Mizushima | Aug 2007 | A1 |
| 20070203944 | Batra et al. | Aug 2007 | A1 |
| 20080044141 | Willis et al. | Feb 2008 | A1 |
| 20080098458 | Smith et al. | Apr 2008 | A2 |
| 20080140844 | Halpern | Jun 2008 | A1 |
| 20080163124 | Bonev et al. | Jul 2008 | A1 |
| 20080195664 | Maharajh et al. | Aug 2008 | A1 |
| 20080286741 | Call | Nov 2008 | A1 |
| 20090019158 | Langen et al. | Jan 2009 | A1 |
| 20090024764 | Atherton et al. | Jan 2009 | A1 |
| 20090034537 | Colrain et al. | Feb 2009 | A1 |
| 20090037367 | Wein | Feb 2009 | A1 |
| 20090150647 | Mejdrich et al. | Jun 2009 | A1 |
| 20090172636 | Griffith et al. | Jul 2009 | A1 |
| 20090182642 | Sundaresan | Jul 2009 | A1 |
| 20090327471 | Astete et al. | Dec 2009 | A1 |
| 20100198920 | Wong et al. | Aug 2010 | A1 |
| 20100199259 | Quinn et al. | Aug 2010 | A1 |
| 20110029812 | Lu et al. | Feb 2011 | A1 |
| 20110055510 | Fritz et al. | Mar 2011 | A1 |
| 20110066737 | Mallart | Mar 2011 | A1 |
| 20110071981 | Ghosh et al. | Mar 2011 | A1 |
| 20110119673 | Bloch et al. | May 2011 | A1 |
| 20110161457 | Sentinelli et al. | Jun 2011 | A1 |
| 20110228668 | Pillai et al. | Sep 2011 | A1 |
| 20110246582 | Dozsa et al. | Oct 2011 | A1 |
| 20120023557 | Bevan et al. | Jan 2012 | A1 |
| 20120218891 | Sundararajan et al. | Aug 2012 | A1 |
| 20130004002 | Duchscher et al. | Jan 2013 | A1 |
| Entry |
|---|
| Richard G. Baldwin, “The ByteBuffer Class in Java : Java Programming Notes #1782”., Aug. 20, 2002. 14 pages. Retrieved from: http://www.developer.com/java/other/article.php/1449271/The-ByteBuffer-Class-in-Java.htm. |
| International Search Report and Written Opinion dated Dec. 6, 2011, International Application No. PCT/US2011/051697, 11 pgs. |
| International Search Report and Written Opinion dated Dec. 6, 2011, International Application No. PCT/US2011/051459, 9 pgs. |
| European Patent Office, International Searching Authority, International Search Report and Written Opinion dated Mar. 5, 2014 for International Application No. PCT/US2012/044429, 18 pages. |
| Oracle International Corporation, Oracle Fusion Middleware—Configuring and Managing JDBC Data Sources for Oracle WebLogic Server 11g Release 1 (10.3.5), Apr. 2011, pp. 1-148. |
| Oracle International Corporation, Oracle Universal Connection Pool for JDBC—Developer's Guide 11g Release 2 (11.2), Sep. 2009, pp. 1-74. |
| Oracle International Corporation, Automatic Workload Management with Oracle Real Application Clusters 11g Release 2, Jan. 2010, pp. 1-31. |
| Oracle International Corporation, Application Failover with Oracle Database 11g, Sep. 2010, pp. 1-5. |
| Pfister, An Introduction to the InfiniBand Architecture, High Performance Mass Storage and Parellel I/O, 2002, pp. 617-632. |
| Number | Date | Country | |
|---|---|---|---|
| 20120066363 A1 | Mar 2012 | US |
| Number | Date | Country | |
|---|---|---|---|
| 61383285 | Sep 2010 | US | |
| 61384227 | Sep 2010 | US |
