Oracle® Fail Safe Concepts and Administration Guide
Release 3.3.1 for Windows
Part No. A96684-01
24 hours a day, 365 days a year.
A cluster configuration in which all cluster nodes perform work. If one node becomes unavailable, one or more other nodes take over the workload of the node that is no longer available.
A cluster configuration in which one node usually stands idle in anticipation of a failover from another node.
The measure of the ability of a system or resource to provide the desired service when required. Availability is measured in terms of the percentage of time the device is accessible out of the total time it is needed. Businesses that require uninterrupted computing services have an availability goal of 24x365.
A protocol that enables clients to retrieve information from an Oracle database without using the network listener. The bequeath protocol internally spawns a server thread for each client application. In a sense, it does the same operation that a remote network listener does for a database connection, but locally.
The application that provides all user-oriented activities, such as character or graphical user display, screen control, data presentation, application flow, and other application-specific tasks.
A group of two or more independent computing systems that operate as a single virtual system.
A node-independent network name that identifies a cluster and is used for cluster-related system management.
A Windows system that is a member of a cluster.
An Oracle Applications process manager that coordinates the processes generated by users' requests to run various data-intensive programs. An Oracle Applications product group can have several concurrent managers. See also internal concurrent manager.
A file that contains the contents of logical database structures, such as tables and indexes. One or more data files form a logical unit of storage called a tablespace. A data file can be associated with only one tablespace, and only one database.
The measure of the inability of a system or resource to provide the desired service when required. Downtime is measured in terms of the percentage or amount of time the device is not accessible out of the total time it is needed.
The process of intentionally returning a group of cluster resources to a preferred owner node from the failover node after the preferred owner node returns to operational status.
The process of taking cluster resources offline on one node and bringing them back online on another node. This process can either be planned (for upgrades and maintenance, for example) or unplanned (due to system or resource failure, for example).
The server node that takes over the workload of an unavailable node.
A user-specified time period in which the cluster software should continue to try to move cluster resources from one node to another before discontinuing the failover process and taking the resources offline. See also group failover policy.
The maximum number of times the cluster software should attempt to move resources from one node to another during the time period (failover period) that you specify. After reaching the specified failover threshold, the cluster software will stop the failover process and take the resources offline. See also group failover policy.
A resource that has been configured for high availability.
A Windows service that is supported by the generic service resource DLL provided with Microsoft Cluster Server (MSCS). The generic service resource DLL is used to configure standard Windows services (such as IP addresses, physical disks, and some applications) as resources in a cluster.
A logical collection of cluster resources that forms a minimal unit of failover. In a failover situation, the group of resources is moved together to a failover node. A group resides on only one cluster node at a time.
A user-specified plan that determines when and if cluster resources should fail back to the preferred owner node from the failover node.
The process of taking a group of cluster resources offline on one node and attempting to bring them back online on another node. This process can either be planned (for upgrades and maintenance, for example) or unplanned (due to system or resource failure, for example).
A user-specified plan that determines two parameters: the time period in which the cluster software should continue to move resources from one node to another (failover period), and the maximum number of times failover should occur during the failover period (failover threshold). See also failover period and failover threshold.
See private interconnect.
A name that represents the specific IP address on a network. In Microsoft Cluster Server (MSCS), the host name is mapped to a network name resource. See also network name.
A combination of System Global Area (SGA) and one or more Oracle database server processes. When a database is started on a database server, Oracle allocates SGA and starts one or more Oracle processes. The memory and processes of an instance efficiently manage the associated database's data and serve the database users. Each instance has a unique Oracle System Identifier (SID), instance name, instance number, rollback segments, and thread ID.
A process manager that monitors requests to run a program, controls other concurrent managers, and determines when a request should be processed and which concurrent manager should carry it out. It is the internal concurrent manager that you configure for high availability using Oracle Fail Safe. See also concurrent manager.
See private interconnect.
The Internet Protocol (IP) address. An IP address takes the form n.n.n.n, for example, 126.96.36.199.
A service that receives requests by clients and redirects them to the appropriate server.
A configuration that includes several Oracle Reports Servers; one is designated as the master; the rest are designated as slaves. Users send report requests to the master Oracle Reports Server. The master Oracle Reports Server then redirects the report requests to the slave servers, using a round-robin algorithm. If one of the slaves is unavailable, the master Oracle Reports Server does not send it requests. The Oracle Reports documentation refers to this as a cluster configuration.
Microsoft Corporation software that provides the capability to cluster individual nodes that are running Windows NT Enterprise Edition, Windows 2000 Advanced Server, or Windows 2000 Datacenter Server software.
A type of business function that is critical to the company and requires high availability.
Network information that describes the network and connection data of an Oracle database. More than one net service name can be defined for an Oracle database. Prior to Oracle8i, a net service name was referred to as a service name entry.
The Microsoft Cluster Server (MSCS) term for a NetBIOS name, which translates into a specific IP address on a network. See also host name.
A computing system that is a member of a cluster.
A configuration in which each node of the cluster performs different work. For example, in a two-node cluster, one node could serve as a database server and another node could serve as an application server.
The process of intentionally taking client applications and cluster resources offline on one node and bringing them back online on another node. For example, the Oracle Fail Safe Installation Guide describes how to perform a planned failover to perform a rolling upgrade (you fail over all resources to one cluster node as you sequentially upgrade software or hardware on another node). See also unplanned group failover.
A node capable of running a specified resource based on the following qualities:
The resource DLL for the specified resource has been installed on the node.
The resource has been configured to run on the node.
You have not manually removed the node from the possible owner nodes list for the resource or the group containing the resource.
In a two-node cluster, both nodes must be possible owner nodes for all resources in a group if you want that group to be able to fail over.
The set of all nodes on which the resource DLL for the specified resource has been installed and configured to run, less any nodes that you explicitly remove from the set.
The node on which you want a group to reside when all cluster nodes that are possible owners are up and running. See also failover node.
In an active/passive configuration, the nodes that perform work. See also active/passive configuration.
A network connection that is dedicated to intracluster communication. The private interconnect is also referred to as a heartbeat connection, because it allows one node to detect the availability or unavailability of another node. The private interconnect is distinct from the public interconnect. See also public interconnect.
A network connection (such as a LAN or WAN) that connects clients to the cluster. See also private interconnect.
A voting mechanism used to guarantee that specific data necessary for recovery can be maintained consistently among all cluster members. This mechanism involves a special storage resource called the quorum resource. The quorum is also used to establish the cluster. See also quorum resource.
The quorum-capable storage resource selected to maintain the configuration data necessary for recovery of the cluster. The quorum resource is generally accessible to other cluster resources so that any cluster node has access to the most recent changes to the configuration data. See also quorum.
Duplicate or extra computing components that safeguard the integrity of a computing system.
A physical or logical component that is available to a computing system. For example, a resource can be a disk, a network IP address, an Oracle database server, a listener, an Oracle Forms Server, an Oracle Reports Server, or a Web server. See also cluster resource and standalone resource.
Relationships between resources in a group that define the order in which the cluster software brings those resources online and offline.
A policy that specifies whether or not a resource failure should result in a group failover.
A policy that specifies whether or not the cluster software should attempt to restart a failed resource on its current node, and if so, how many attempts within a given time period should be made to restart it.
A software installation technique that allows a cluster system to continue to provide service while the software is being upgraded to the next release. This process is called a rolling upgrade because each node is upgraded and rebooted in turn, until all cluster systems and client nodes have been upgraded. While each node is temporarily offline, another node takes over the workload of the node being upgraded.
An optional, preconfigured starter database that is provided with Oracle Fail Safe so you can try out the functions of Oracle Fail Safe before using them on your production database.
See net service name.
A cluster configuration in which all cluster nodes are cabled physically to the same disks, but only one node can access a given disk at a time for either read or write activity.
An I/O connection on which the cluster disks are accessible from all nodes in a cluster.
An installation method that allows you to install software by supplying input to Oracle Universal Installer with a response file.
A 32-bit value that indicates how many bits in an address are being used for the network ID.
The ability of client applications to automatically reconnect to a database and resume work after a failover occurs.
A software-initiated failover process that is triggered automatically in response to a software or hardware failure. See also planned group failover.
A network address at which resources in a group can be accessed, regardless of the cluster node hosting those resources. A virtual address on an MSCS cluster consists of a network name and associated IP address.
A name that maps to a physical directory specification. You specify a virtual directory to hide your file structure from users. If the physical directory changes, the URL specified by users does not change. For example, if your virtual address is "Company," and you have mapped the virtual directory "Sales" to U:\SalesInfo\Webfiles, users will access sales information by entering the URL http://Company/Sales.
A group with one or more virtual addresses.