|Oracle9i Real Application Clusters Concepts
Release 2 (9.2)
Part Number A96597-01
A software synchronization feature used for notification. Real Application Clusters also uses a blocking interrupt.
In a client/server model, the part of the system that performs information preparation and exchange on behalf of a client or server. In the phrase, intelligent agent, it implies an automatic process that can communicate with other agents to perform some collective tasks on behalf of one or more users.
A software synchronization feature that notifies the process holding the access right to a resource that another process needs to access the same resource in an incompatible mode. (The shared (S) mode and exclusive (X) mode, for example, are incompatible). See also acquisition interrupt.
The state of a buffer in the local cache of an instance.
The synchronization of data in multiple caches so that reading a memory location through any cache will return the most recent data written to that location through any other cache. Sometimes called cache consistency.
A diskless cache coherency mechanism in Real Application Clusters that provides copies of blocks directly from a holding instance's memory cache to a requesting instance's memory cache.
A set of instances that cooperates to perform the same task.
The generic term for a Real Application Clusters database.
See cluster database.
See cluster database.
An operating system-dependent component that discovers and tracks the membership state of each node by providing a common view of cluster membership across the cluster. The CM also monitors process health, specifically the health of the database instance. The Global Enqueue Service Monitor (LMON) process, a background process that monitors the health of the Global Cache Service (GCS), registers and de-registers from the CM.
The Global Cache Service (GCS) ensures that a consistent read block (also known as the master copy data block) is maintained. The consistent read block is the master block version that records information about all changes to a block. It is held in at least one System Global Area (SGA) in the cluster if the block is to be changed. If an instance needs to read the block, then the current version of the block may reside in many buffer caches as a shared resource. Thus, the most recent copy of the block in all System Global Areas contains all changes made to that block by all instances, regardless of whether transactions on those instances have committed.
An operating system running a program runs in either user mode or operating system mode. Switching between user and operating system mode is a context switch. For example, a program that makes a system call while in user mode makes a context switch. Context switches can hinder performance because, in this example, the context of the user program must be stored while also transferring the context of the operating system kernel into memory. Performance can be even more adversely affected when multiple system calls compete for operating system resources.
A portion of a physical disk where an extended partition is created, logical partitions are assigned, and formatting has been completed. In contrast, an unformatted partition is called a raw partition.
See consistent read (CR).
Abbreviation for disk and execution monitor. A program that is not invoked explicitly, but lies dormant waiting for specific conditions to occur.
An Oracle tool for creating and deleting databases and for managing database templates.
See Oracle9i Data Guard.
Database and application environments that help with decision support or data warehouse systems.
A Real Application Clusters background process that captures diagnostic data on instance process failures. No user control is required for this daemon.
A data block that has been changed by an instance.
Shared memory structures that serialize access to database resources. Enqueues are local to one instance if Real Application Clusters is not enabled. When you enable Real Application Clusters, enqueues can be local or global to a database. (See also: latch, lock, and resource.)
A tool for creating, deleting, and modifying Enterprise Manager configurations and settings.
The buffer state name for an exclusive resource.
The means of failure recognition and recovery used by Real Application Clusters.
The ability of a system or component to continue normal operation despite the presence of hardware or software faults. This normally involves some degree of redundancy.
A standard of the American National Standards Institute (ANSI) for a 100 Mb per second local area network architecture. The underlying medium is often optical fiber and the topology is a dual-attached, counter-rotating token ring.
The generic term for a high speed serial data transfer architecture recently standardized by the American National Standards Institute (ANSI). The Fibre Channel architecture was developed by the Fibre Channel Industry Association (FCIA), a consortium of computer and mass storage manufacturers. The best-known Fibre Channel standard is the Fibre Channel Arbitrated Loop (FC-AL). (See also: Storage Area Network (SAN).
A colloquial term for 99.999% system availability.
In Real Application Clusters, a particular data block can only be modified by one instance at a time. If one instance modifies a data block that another instance needs, then whether a forced disk write is required depends on the type of request submitted for the block.
A set of free lists available for use by one or more instances.
An Oracle-specific data structure representing a Cache Fusion resource. There is a 1:1 corresponding relationship between a global cache element and a Cache Fusion resource in the Global Cache Service (GCS).
The process that implements Cache Fusion. It maintains the block mode for blocks in the global role. It is responsible for block transfers between instances. The Global Cache Service employs various background processes such as the Global Cache Service Processes (LMSn) and Global Enqueue Service Daemon (LMD).
The processes that handle remote Global Cache Service (GCS) messages. Real Application Clusters provides for up to 10 Global Cache Service Processes. The number of LMSn varies depending on the amount of messaging traffic amount the nodes in the cluster. The LMSn handle the acquisition interrupt and blocking interrupt requests from a remote instance for Global Cache Service resources. For cross-instance consistent read requests, LMSn creates a consistent read version of the block and sends it to the requesting instance. LMSn also controls the flow of messages to and from remote instances.
This service coordinates local and global enqueues.
The resource agent process that manages Global Enqueue Service (GES) resource requests. The LMD process also handles deadlock detection Global Enqueue Service (GES) requests. Remote resource requests are requests originating from another instance. (See also: daemon.)
The process that monitors the entire cluster to manage global resources. LMON manages instance and process expirations and the associated recovery for the Global Cache Service (GCS) and Global Enqueue Service (GES). In particular, LMON handles the part of recovery associated with global resources.
The data structures associated with global resources. It is distributed across all instances in a cluster.
A component that receives requests from SRVCTL to execute administrative job tasks, such as startup or shutdown. The command is executed locally on each node, and the results are sent back to SRVCTL. The daemon by default. It should not be deleted.
See high availability.
A type of failover performed by the platform-specific cluster manager (CM). If a node or the instance running on it fails, the cluster manager restarts the instance on another node in the cluster. Restarting the Oracle instance requires moving the IP addresses, volumes, and file systems containing the Oracle datafiles. It also requires starting the Oracle server and opening the datafiles on the new node.
A periodic message that shows that an instance is active.
A system type with redundant components that provides consistent and uninterrupted service, even in the event of hardware or software failures.
An initialization parameter file that contains global parameters that apply to an entire cluster.
The combination of the System Global Area (SGA) and each process for the Oracle database. The memory and processes of an instance manage the associated database's data and serve the database users. Each instance has unique system identifier (SID), instance name, rollback segments, and thread ID.
The method used by Real Application Clusters guaranteeing that all cluster members are functional. IMR polls and arbitrates the membership. Any members that do not show a heartbeat in the control file record or who do not respond to periodic status queries are presumed to have expired. The Instance Membership Recovery arbiter settles the membership votes of nodes in a cluster, and creates a control file voting results record (CFVRR).
Represents the name of the instance and is used to uniquely identify a specific instance when clusters share common services names. The instance name is identified by the INSTANCE_NAME parameter in the instance initialization file, initsid.ora. The instance name is the same as the Oracle system identifier (sid). The instance name is the same as the system identifier (SID).
The communication link between the nodes.
A high-speed operating system-dependent transport component. The IPC transfers messages between instances on different nodes. Also referred to as the interconnect.
In the context of databases, the set of foreground and background processes that implement a database.
A simple, low-level serialization mechanism that protect in-memory data structures in the System Global Area (SGA). Latches do not protect datafiles, are automatic, and are held for a very short time in exclusive mode. Because latches are synchronized within a node, they do not facilitate internode synchronization. (See also: enqueue, lock, and resource.)
A process that manages instance global enqueue requests and cross-instance call operations. Workload is automatically shared and balanced when there are multiple Global Cache Service Processes (LMSn).
Objects in the library cache can become invalid due to object dependencies. The object is recompiled on its next use.
A process that resides on the server to listen for incoming client connection requests and manage the traffic to the server. When a client requests a network session with a server a listener receives the request. If the client information matches the listener information, then the listener grants a connection to the server.
A listener configuration file that identifies the protocol addresses on which the listener is accepting connection requests and the services the listener listens for.
Synchronization mechanisms that coordinate concurrent access to shared data structures in cluster databases.
A list of blocks containing available space from any extent in a table.
See resource mastering.
The average time (usually expressed in hours) that a component works without failure. It is calculated by dividing the total number of failures into the total number of operating hours observed. The term can also mean the length of time a user can reasonably expect a device or system to work before an failure occurs.
See null (N) mode.
A node is a machine on which an instance resides.
The processing of transactions by computers in real time. (See also: transaction systems.)
Clusterware tailored for various operating systems. OSD clusterware provides communication links between the operating system and the Real Application Clusters software.
A system management tool that provides an integrated solution for centrally managing your heterogeneous environment. Oracle Enterprise Manager combines a graphical console, management server, Oracle Intelligent Agent, repository database, and tools to provide an integrated, comprehensive systems management platform for managing Oracle products.
A software component that enables connectivity. It includes a core communication layer called the Oracle Net foundation layer and network protocol support. Oracle Net enables services and their applications to reside on different computers and communicate as peer applications.
A post-installation tool that configures basic network components after installation.
The term that encompasses all of the Oracle networking components, including: Oracle Net, the listener, Oracle Connection Manager, Oracle Names, Oracle Net Configuration Assistant, and Oracle Net Manager.
An add-on application for Oracle Enterprise Manager that offers a variety of tabular and graphic performance statistics for Real Application Clusters. The statistics represent the aggregate performance for all instances running on Real Application Clusters.
A breakthough architecture that enables clusters to access a shared database. Real Application Clusters includes the software component that provides the necessary Real Application Clusters scripts, initialization files, and datafiles to make the Oracle9i Enterprise Edition an Oracle9i Real Application Clusters database.
A failover protection feature. Oracle Real Application Clusters Guard is an integral component of Real Application Clusters. Oracle Real Application Clusters Guard provides the following functions:
Real Application Clusters Guard II enables you to monitor clusters that use Real Application Clusters to maintain availability. On failure, Real Application Clusters Guard II optimally transfers application service loads to other active nodes. In Real Application Clusters Guard II, all instances are active and are able to support services. Instances are brought in automatically to support a service while they are also available to support other services.
Divides the work of processing certain types of SQL statements among multiple parallel execution server processes. Commonly used in Decision Support System (DSS) applications.
A file used by an Oracle server that provides specific values and configuration settings that are used at database startup. The keyword
PFILE is used in the startup command.
A copy of a dirty block that is used by the Global Cache Service (GCS). Past images of blocks are maintained until writes covering those versions are recorded. Past images are used in failure recovery.
See past image (PI).
A synonymous term for forced disk write.
Includes routine operations, maintenance, and upgrades that cause the system to be unavailable to users. (See also: unplanned downtime.)
In primary/secondary configurations, the instance that mounts the database first assumes the primary role. It performs the work requested by application sessions. If the primary instance fails or is shut down, then failover occurs, and another instance assumes the primary instance role. (See also: Primary/Secondary Instance Configuration and secondary instance role.)
A configuration in which the primary instance is the instance where all clients access the database. The secondary instance provides backup services to the primary instance in case the primary instance fails. (See also: primary instance, primary instance role, secondary instance, and secondary instance role.)
A rollback segment that is acquired exclusively by an instance when the instance opens a database.
A rollback segment that is available to any instance that requires a rollback segment.
A disk drive that does not yet have a file system set up. Raw devices are used for Real Application Clusters since they enable the sharing of disks. See also raw partition.
A portion of a physical disk that is accessed at the lowest possible level. A raw partition is created when an extended partition is created and logical partitions are assigned to it without any formatting. Once formatting is complete, it is called a cooked partition. See also raw device.
See raw device.
An Oracle tool that enables you to back up, copy, restore, and recover datafiles, control files, and archived redo logs. It is included with the Oracle server and does not require separate installation.
A hardware architecture that combines multiple hard disk drives to allow rapid access to a large volume of stored data.
A set of tables in an Oracle database that stores data required by Oracle Enterprise Manager. This database is separate from the database on the nodes.
A concurrency control that defines global access rights for instances in a cluster.
A concurrency control that defines whether a data block is cached in only one instance (local) or if it cached in multiple instances (global).
The memory that stores recently accessed data for an individual record (or row). Row caches are used so that subsequent requests for the data held in the same row can be processed more quickly.
See shared (S) mode.
The ability to add additional nodes to Real Application Clusters environments and maintain and often achieve markedly improved performance.
The factor that expresses how much more work can be done in the same time period by a larger system. the factor that expresses how much more work can be done in the same time period by a larger system. (See also: speed up.)
In a Primary/Secondary Instance Configuration, the second instance to mount the database assumes the secondary role. The instance with the primary role performs the work that is requested by application sessions, but selected tasks such as reporting and planned operations can be performed by the instance with the secondary instance role. (See also: primary instance, primary instance role, and secondary instance.)
An enqueue utilized by a session when it needs a sequence value.
A logical representation of a database. This is the way a database is presented to clients. A database can be presented as multiple services and a service can be implemented as multiple database instances. The service name is a string that includes:
The service name is entered during installation or database creation.
If you are not sure what the global database name is, you can obtain it from the combined values of the
SERVICE_NAMES parameter in the common database initialization file, initdbname.ora.
The service name is included in the
CONNECT_DATA part of the connect descriptor.
A feature whereby PMON, the process monitor, automatically registers information with a listener. Because this information is registered with the listener, the
listener.ora file does not need to be configured with this static information.
The aggregation of the buffer caches of all instances in a cluster database.
The buffer state name for a shared resource.
A protected read resource mode. No writes are allowed in shared mode. In shared mode, any number of users can have simultaneous read access to a resource. (See also: exclusive (X) mode and null (N) mode.)
A server that is configured to allow many user processes to share very few server processes, so that the number of users that can be supported is increased. With shared server configuration, many user processes connect to a dispatcher.
The concept that more hardware can perform the same task in less time than the original system. With added hardware, speed up holds the task constant and measures time savings. (See also: scale up.)
The Service Control (SRVCTL) utility that administrators can use to manage instances and Real Application Clusters databases. SRVCTL is installed on each node. The SRVCTL utility gathers information about all the instances for Oracle Enterprise Manager. SRVCTL acts as a single point of control between the Oracle Intelligent Agent and the nodes. Only one node's Oracle Intelligent Agent is used to communicate to SRVCTL. SRVCTL on that node then communicates to the other nodes through Java Remote Method Invocation (RMI).
See Oracle9i Data Guard.
A high speed network of shared storage devices. Typically, SAN architecture enables access to all storage devices by all servers on a local area network (LAN) or wide are network (WAN). Most SANs conform to Fibre Channel standards.
Two or more similar processors connected through a high-bandwidth link and managed by one operating system, where each processor has equal access to I/O devices.
The Oracle system identifier (SID) identifies a specific instance of the running Oracle software. For a Real Application Clusters database, each node within the cluster has an instance referencing the database.
The database name, specified by the
DB_NAME parameter in the
INITDB_NAME.ORA file, and unique thread ID make up each node's SID. The thread ID starts at 1 for the first instance in the cluster, and is incremented by 1 for the next instance, and so on.
Each Oracle instance has its own set of online redo log groups. These groups are called a thread of online redo. In non-Real Application Clusters environments, each database has only one thread that belongs to the instance accessing it. In Real Application Clusters environments, each instance has a separate thread, that is, each instance has its own online redo log. Each thread has its own current log member.
A colloquial term for 99.9% system availability.
A file that contains net service names. This file is needed on clients, nodes, the Console, and the Oracle Performance Manager machine.
A list of blocks freed by uncommitted transactions.
Application systems characterized by updates to the database. Examples of transaction systems include e-business systems and ERP applications. More specialized examples include telephone call and billing systems, credit card transactions, and airline reservation systems. Transaction systems are also known as Online Transaction Processing (OLTP) systems.
An action is transparent if it takes place without any effect that is visible to users.
A runtime failover for high availability environments, such as Real Application Clusters and Oracle Real Application Clusters Guard I. TAF refers to the failover and reestablishment of application-to-service connections. It enables client applications to automatically reconnect to the database if the connection fails, and optionally resume a
SELECT statement that was in progress. This reconnect happens automatically from within the Oracle Call Interface (OCI) library.
System downtime that includes system faults, data and media errors, and site outages that cause the system to be unavailable to users. (See also: planned downtime.)
Software that provides storage management for a Storage Area Network (SAN). Usually host-based Redundant Array of Independent Disks (RAID) software. Most volume managers use a graphical user interface (GUI) console to show the partitioning and status of storage volumes. Volume managers can create volumes from various disks, and those volumes can encompass multiple disks. (See also: raw partition.)
The process of transferring information about parsed SQL statements and compiled PL/SQL units from the library cache on the primary instance to the library cache on the secondary instance. Warming the cache improves performance after failover because the library cache is already populated.
See exclusive (X) mode.