Oracle Automatic Storage Management (ASM) simplifies the management of storage resources in Oracle RAC environments. ASM provides an optimized, high-performance storage solution, and its tight integration with Oracle RAC ensures that data is distributed across all available disks, providing fault tolerance and improving performance. This chapter will explain how ASM works in Oracle RAC, discuss ASM disk groups, redundancy levels, failure groups, and provide step-by-step commands for installing, configuring, and managing ASM.

---

1. ASM and Its Role in Oracle RAC

ASM is Oracle’s volume manager and file system designed specifically for Oracle database files. It provides the following benefits:

• Simplifies storage management: ASM handles striping, redundancy, and rebalance automatically.
• High availability and performance: ASM distributes data evenly across all available disks, providing load balancing and better performance.
• Fault tolerance: By creating disk groups with redundancy, ASM ensures high availability in case of disk failures.
• Integrated with Oracle RAC: ASM manages shared storage, making it ideal for Oracle RAC environments where multiple instances access the same data.

---

2. ASM Disk Groups, Redundancy Levels, and Failure Groups

ASM Disk Groups:

ASM disk groups are collections of disks managed as a single logical unit. All database files, including datafiles, control files, and redo logs, reside in disk groups. Each disk group can consist of multiple physical disks, and ASM automatically stripes data across these disks.

Redundancy Levels in ASM:

1. External Redundancy:
   • No data mirroring is provided by ASM. This redundancy level assumes that the underlying storage provides redundancy (e.g., RAID).
   • Suitable for environments where hardware-level redundancy is present.
2. Normal Redundancy:
   • ASM provides two-way mirroring. Each file is mirrored across two disks.
   • Provides protection against single disk failures.
3. High Redundancy:
   • ASM provides three-way mirroring, ensuring data is replicated across three disks.
   • Provides the highest level of protection but consumes more storage space.

Failure Groups:

Failure groups are a logical grouping of disks within a disk group, where disks in a failure group share a common failure characteristic (e.g., disks connected to the same storage controller). ASM ensures that mirrored copies of data are stored in different failure groups to prevent data loss in case a failure group goes down.

Example Commands for Managing ASM Disk Groups:

• Create an ASM disk group (with external redundancy):

asmcmd createdg DATA external /dev/sdb /dev/sdc

Create an ASM disk group (with normal redundancy):

asmcmd createdg DATA normal /dev/sdb /dev/sdc /dev/sdd

List ASM disk groups:

asmcmd lsdg

Add a new disk to an existing disk group:

asmcmd adddisk DATA /dev/sde

Remove a disk from a disk group:

asmcmd rmdisk DATA /dev/sdb

3. Installing and Configuring ASM

To install and configure ASM in Oracle RAC, you need to install Oracle Grid Infrastructure. This installation includes both Oracle Clusterware and ASM. ASM instances are created automatically on each node during Grid Infrastructure installation.

Steps for Installing ASM:

1. Pre-Installation: Ensure that all necessary Linux kernel parameters are set, and ASM disks are available for use by Oracle Grid Infrastructure.Example of setting kernel parameters:

sysctl -w kernel.shmmax=4294967295
sysctl -w kernel.sem="250 32000 100 128"

Prepare ASM disks: Identify the disks that will be used for ASM. Use fdisk to create partitions or configure them as raw devices.

# List all available disks
fdisk -l

# Partition a disk for ASM
fdisk /dev/sdb

Install Oracle Grid Infrastructure: Run the Oracle Universal Installer (OUI) to install Oracle Grid Infrastructure, which includes ASM.

OUI Installation Steps:

• Select “Install and Configure Grid Infrastructure for a Cluster”.
• Choose “Configure ASM” during installation.
• Specify the disk groups and redundancy level for ASM.

Post-Installation Configuration: After installation, ensure ASM is properly configured.

Start the ASM instance:

srvctl start asm

Verify the ASM instance is running:

srvctl status asm

4. Managing ASM Instances with Oracle RAC

In Oracle RAC, each node has its own ASM instance, which is responsible for managing the shared storage. The ASM instances work together to provide a single, coherent view of the storage resources to the database instances.

Key ASM Management Commands:

• Start and Stop ASM Instances: Start ASM on all nodes:

srvctl start asm -n all

Stop ASM on all nodes:

srvctl stop asm -n all

Check the Status of ASM Instances:

srvctl status asm

Mount an ASM Disk Group:

srvctl start diskgroup -g DATA

Dismount an ASM Disk Group:

srvctl stop diskgroup -g DATA

Monitoring ASM:

• Check ASM instance parameters:

sqlplus / as sysasm
SQL> show parameter asm;

List all disks in ASM:

asmcmd lsdsk

Check disk group usage:

asmcmd du

Monitor ASM rebalancing operations:

asmcmd lsop

5. Best Practices for Using ASM in a RAC Environment

Use Separate Disk Groups for Different Types of Data:

• It’s recommended to create separate disk groups for data, logs, and backups. This helps optimize performance and manage storage more effectively.

Example:

asmcmd createdg DATA normal /dev/sdb /dev/sdc
asmcmd createdg FRA external /dev/sdd /dev/sde

Ensure Disk Group Redundancy Matches the Use Case:

• Use normal redundancy for critical data where disk failures are expected, and external redundancy when redundancy is already handled at the hardware level (e.g., RAID).

1. Balance Load Across Disks:
   • When adding disks to a disk group, ASM automatically rebalances the data across all available disks. However, it’s important to monitor disk utilization and ensure that the rebalancing is occurring as expected.
   Monitor Rebalancing:

asmcmd lsop

Enable ASM Disk Scrubbing:

• Disk scrubbing checks the consistency of the data on the disks and can prevent corruption over time. It’s important to enable ASM scrubbing for all critical disk groups.

Enable scrubbing:

sqlplus / as sysasm
SQL> alter diskgroup DATA scrub power low;

Monitor ASM Alerts and Logs:

• Regularly check the ASM alert logs for any warnings or errors that may affect the performance or availability of the disk groups.

View the ASM alert log:

tail -f /u01/app/grid/diag/asm/+asm/trace/alert_+ASM1.log

Conclusion

In this chapter, we have explored the role of Automatic Storage Management (ASM) in Oracle RAC environments, focusing on its ability to simplify storage management, improve performance, and provide fault tolerance. We covered ASM disk groups, redundancy levels, failure groups, and provided step-by-step instructions on installing, configuring, and managing ASM in a RAC environment. Following best practices for using ASM will help ensure high availability and optimal performance in Oracle RAC deployments.