Oracle RAC Components

Oracle RAC (Real Application Clusters) comprises several key components that work together to provide high availability, scalability, and fault tolerance. Below is a detailed explanation of these components with examples and associated commands.

---

1. Cluster Nodes:

A cluster node is a server that is part of the Oracle RAC environment. All the nodes in the RAC cluster work together and provide access to the same shared database. Each node runs an Oracle database instance, and all instances access the same set of data files on shared disk storage.

• Example: In a 3-node RAC setup, if Node 1 fails, Nodes 2 and 3 will continue to serve database requests, ensuring no downtime for the users.
• Key Command:

crsctl status resource -t

This command shows the status of the cluster resources, including nodes and database instances, in an Oracle RAC environment.

2. Shared Disk:

In Oracle RAC, all cluster nodes must be able to access the same shared storage. This allows each instance to read and write to the same database files. Shared disk storage is managed by ASM (Automatic Storage Management) in most RAC configurations, which optimizes disk performance and redundancy.

• Example: A 2-node RAC cluster is connected to shared disk storage through SAN (Storage Area Network) or NAS (Network Attached Storage). Both nodes can access and update the same data stored on shared disks.
• Key Command:

SELECT * FROM v$asm_diskgroup;

This SQL query retrieves information about the ASM disk groups, showing the shared storage used by RAC instances.

3. Oracle Clusterware:

Oracle Clusterware is the software layer that enables servers (nodes) to work together in a cluster. It manages node membership, cluster resources, and failover services, ensuring the availability of the RAC environment. It is responsible for detecting and resolving failures of nodes or other cluster components.

• Example: If one node in the RAC environment goes down, Oracle Clusterware will automatically failover its services to another node, ensuring continuous availability.
• Key Command:

crsctl query crs activeversion

This command checks the version of the active Oracle Clusterware running on your RAC nodes.

Global Cache and Cache Fusion:

Oracle RAC uses Global Cache to coordinate access to the same data blocks by multiple nodes. Oracle’s Cache Fusion technology ensures that data changes are shared in real time between instances without requiring disk I/O. Instead of writing changes to disk, Cache Fusion uses the high-speed interconnect to pass modified data between nodes.

• Example: If Node 1 updates a data block in memory, Node 2 will request the updated version of that data block via Cache Fusion. This ensures that Node 2 gets the most recent version without needing to read the block from disk.
• Key Command:

SELECT event, total_waits FROM v$system_event WHERE event LIKE 'global cache%';

This SQL query shows statistics about global cache events, helping you monitor the behavior of Cache Fusion in the RAC environment.

Oracle RAC Processes:

Oracle RAC has several background processes that handle coordination and communication between instances in the cluster. The most important ones include:

1. GES (Global Enqueue Service):
   Manages locks and resources across the cluster and ensures consistency when multiple instances access shared resources.
2. GCS (Global Cache Service):
   Ensures data block consistency across instances. It tracks the ownership of data blocks and manages block transfers between instances in Cache Fusion.
3. LMS (Lock Manager Server):
   Responsible for handling requests for global cache access and managing lock contention between instances.
4. LMD (Lock Manager Daemon):
   Handles global enqueue requests, converting and granting locks across the RAC nodes.

• Example: In a 3-node RAC cluster, if Node 1 needs a data block held by Node 2, the LMS process on Node 1 will communicate with LMS on Node 2 to obtain the block.
• Key Command:

SELECT name, value FROM v$sysstat WHERE name LIKE 'global%';

This query provides statistics on global cache and lock management activities, which can help diagnose RAC performance issues.

Role of Oracle Grid Infrastructure:

Oracle Grid Infrastructure is the foundation for Oracle RAC and ASM. It includes Oracle Clusterware and ASM to provide all the services necessary for managing cluster databases, nodes, and shared storage. Grid Infrastructure manages both the RAC cluster and ASM disk groups, making it a central component of Oracle’s high-availability architecture.

• Key Features:
  • Cluster management: Manages the membership of cluster nodes.
  • Automatic failover: Detects node failures and automatically redistributes services.
  • ASM management: Manages disk storage for databases.
• Key Command:

crsctl start cluster

• This command starts the Oracle Clusterware, enabling the nodes to function as a cohesive RAC system.

---

Understanding RAC and ASM (Automatic Storage Management):

ASM is Oracle’s storage management solution that is used to manage shared storage in RAC environments. ASM simplifies disk management by abstracting the physical storage into disk groups and providing automated data balancing and redundancy.

• Key Features of ASM:
  • Disk Group: ASM organizes disks into disk groups, where the database files are stored.
  • Automatic Rebalancing: ASM automatically rebalances data when disks are added or removed.
  • Mirroring: ASM provides redundancy through mirroring (normal and high redundancy options).
• Example: In a 2-node RAC setup, both nodes use ASM for managing storage. If a new disk is added, ASM automatically redistributes the data across all available disks to optimize performance and storage utilization.
• Key Commands:
  • To create a disk group:

CREATE DISKGROUP data_disk_group NORMAL REDUNDANCY DISK '/dev/disk1', '/dev/disk2';

To add a disk to a disk group:

ALTER DISKGROUP data_disk_group ADD DISK '/dev/disk3';

To monitor ASM rebalancing:

SELECT * FROM v$asm_operation;

Conclusion:

Oracle RAC provides a robust, highly available, and scalable architecture through its key components like cluster nodes, shared disk, and Oracle Clusterware. Global Cache and Cache Fusion ensure that data consistency is maintained across instances, while background processes like GES, GCS, LMS, and LMD manage the coordination of resources and cache. Oracle Grid Infrastructure ties everything together, offering cluster management, storage solutions via ASM, and automated failover for mission-critical applications. By mastering these components and processes, DBAs can effectively manage RAC environments to ensure optimal performance and availability.