- 25 Jul, 2025
- 0 Comments
- 4 Mins Read
Oracle Clusterware Architecture
Overview of Oracle Clusterware
Oracle Clusterware is a software that provides clustering services and high availability for Oracle databases. It allows multiple servers to work together as a single system, sharing resources to increase performance, scalability, and availability
Oracle Clusterware manages the following
- Cluster Membership: Determines which nodes are part of the cluster and controls the addition or removal of nodes.
- Resource Management: Manages and monitors cluster resources like databases, services, and applications.
- Event Management: Detects and reacts to failures, ensuring failover and recovery of resources.
Key Components of Oracle Clusterware
Oracle Clusterware consists of several key components:
1 . Voting Disk
- The voting disk is used to manage cluster membership by recording which nodes are active members of the cluster.
- Each node in the cluster must be able to access the voting disk. If a node loses access to the voting disk, it is evicted from the cluster.
- Multiple voting disks can be configured for redundancy. Oracle recommends using an odd number of voting disks to avoid split-brain scenarios.
2 . Oracle Cluster Registry (OCR)
- The OCR stores configuration information about the cluster and the resources it manages, such as information about the nodes, databases, and services.
- The OCR is critical for the operation of Oracle Clusterware and must be accessible for the cluster to function.
- OCR is mirrored for redundancy, ensuring high availability.
3. Cluster Synchronization Services (CSS)
- CSS manages cluster node membership and provides messaging among nodes.
- It ensures that all nodes have a consistent view of which nodes are part of the cluster.
- CSS also coordinates the fencing of nodes that are no longer able to participate in the cluster due to network partitioning or failures.
4 . Oracle High Availability Services (OHAS)
- OHAS is responsible for starting, stopping, and monitoring Oracle Clusterware components on each node.
- It manages processes like the Cluster Ready Services (CRS), Event Management (EVM), and the OCR process.
5 . Cluster Ready Services (CRS)
- CRS is responsible for managing Oracle resources, including databases, listeners, and VIPs (Virtual IP addresses).
- CRS monitors the state of resources and ensures that they are running as expected.
- If a resource fails, CRS attempts to restart it on the same node or relocate it to another node in the cluster.
6 . Event Management (EVM)
- EVM provides a framework for managing events within the cluster.
- It captures events like node failures, resource status changes, and system state changes, allowing administrators to respond to these events appropriately.
7 . Oracle Local Registry (OLR)
- The OLR is a node-local repository that contains information about Oracle Clusterware that is specific to the local node.
- It is used during the startup and shutdown of Oracle Clusterware before the OCR is available.
Networking Components
Oracle Clusterware heavily depends on network communication for cluster coordination. There are two main network interfaces used in a typical Oracle Clusterware setup:
1 . Public Network
- The public network is used for client connections to the databases running in the cluster. It provides connectivity for user applications and Oracle RAC services.
2 . Private Network (Interconnect)
- The private network, also known as the interconnect, is used for internode communication within the cluster.
- It carries critical cluster heartbeat information and Oracle RAC cache fusion traffic, making its performance and reliability crucial.
3 . Fencing and Node Eviction
- Fencing (or node eviction) is the process of isolating a node from the cluster when it is no longer able to participate, typically due to network or hardware issues.
- Oracle Clusterware uses a voting mechanism to determine which nodes should be evicted. If a node cannot access the majority of the voting disks, it will be fenced off to protect the integrity of the cluster.
4 . Failover and High Availability
Oracle Clusterware ensures high availability by detecting failures of cluster nodes and automatically failing over services and resources to surviving nodes. This prevents downtime for end-users and maintains the overall performance and reliability of the database system.
- Automatic Restart: If a node fails, Oracle Clusterware will automatically restart the resources on another node.
- Failover: In case of node failure, Oracle Clusterware can automatically relocate the VIP and Oracle RAC instances to another available node.
5 . Integration with Oracle RAC
Oracle Clusterware is tightly integrated with Oracle RAC, providing the infrastructure for Oracle RAC to operate as a multi-node database system. Oracle RAC relies on Oracle Clusterware for node membership management, internode communication, and failover capabilities.
6 . Cluster Verification Utility (CVU)
Oracle provides the Cluster Verification Utility (CVU) to assist with verifying the integrity of the cluster. CVU can be used to validate the cluster configuration, network setup, storage, and other essential components.
7 . Oracle ASM and Oracle Clusterware
Oracle Clusterware integrates with Oracle Automatic Storage Management (ASM), which manages the shared storage for Oracle RAC. Oracle ASM provides high availability and redundancy for database storage, working hand-in-hand with Clusterware to ensure data integrity and availability.
8 . Automatic Diagnostic Repository (ADR)
Oracle Clusterware uses ADR for diagnostics. It stores diagnostic data like log files, trace files, and incident reports, making it easier to diagnose issues within the cluster.
9 . Additional High-Availability Features
- Oracle Grid Infrastructure: Provides a complete stack for high availability, including Oracle Clusterware and ASM.
- Extended Clusters: Oracle Clusterware supports extended clusters, which span multiple geographic locations for disaster recovery.