My Oracle Support Banner

IBM General Parallel File System (GPFS) 3.1 and lower and Oracle RAC (Doc ID 302806.1)

Last updated on MAY 20, 2021

Applies to:

Oracle Database - Enterprise Edition
IBM AIX on POWER Systems (64-bit)
IBM AIX Based Systems (64-bit)



Purpose

Refer to <note 1376369.1> for latest certification information.

 

Overview.

GPFS has been verified for use with:

Scope

Software support details can be found in the table GPFS and Oracle RAC - Supported Software Configurations below:

GPFS is IBM’s high-performance parallel, scalable file system for IBM UNIX clusters capable of supporting multi-terabytes of storage within a single file system.

GPFS is a shared-disk file system where every cluster node can have parallel, concurrent read/write access to the same file.It is designed to provide high-performance I/O by "striping" data across multiple disks accessed from multiple servers.GPFS provides high availability through logging and replication, and can be configured for automatic failover from both disk and node malfunctions.

GPFS 2.1 and 2.2 were previously approved for Oracle RAC but GPFS 2.3 onwards now offers several new key features including:

A summary of key GPFS features for Oracle RAC Databases is given below.  All Oracle RAC configurations planning to use GPFS should select GPFS 3.1 or 3.2 while existing GPFS and Oracle RAC users should consider upgrading to GPFS 2.3.

This document also includes:

Details

To view full details, sign in with your My Oracle Support account.

Don't have a My Oracle Support account? Click to get started!


In this Document
Purpose
 Overview.
Scope
Details
 GPFS and Oracle RAC - Supported Software Configurations:
 Key GPFS features for Oracle RAC Databases on AIX 5L and Linux on POWER.
 GPFS Tuning Requirements for Oracle.
 AIO and DIO options:
 Configuring LUNS for GPFS and Oracle:
 GPFS Block Size, Oracle "db_block_size" and "db_file_multiblock_read_count":
 GPFS and AIX 5L Tuning for AIO
 GPFS Threads
 Corresponding tuning of AIX AIO maxservers.
 GPFS and pinned SGA
 Other important GPFS files system attributes:
 GPFS 2.3 Installation Examples:
 Example 1: Creating A GPFS file system without replication and with tie-breaker disks that are part of the GPFS file system.
 1.      Pre-installation steps.
 2.      Install GPFS software.
 3.      Create a 2 node GPFS cluster.
 4.      Create the cluster-wide names for the Network Shared Disks (NSDs) to be used by GPFS.
 5.      Further customize the cluster configuration and designate tie-breaker disks.
 6.      Start GPFS on all nodes:
 7.      Create and mount the GPFS file system:
 Example 2: Creating A GPFS file system using GPFS replication and tie-breaker disks that are not part of the GPFS file system.
 1.      Pre-installation steps.
 2.      Install GPFS software.
 3.      Create a 2 node GPFS cluster.
 4.      Create the cluster-wide names for the Network Shared Disks (NSDs) to be used for the GPFS file system.
 5.      Create the cluster-wide names for the Network Shared Disks (NSDs) to be used for the tie-breaker disks.
 6.      Identify the tie breaker disks to the cluster configuration using the ‘mmchconfig’ command.
 7.      Start GPFS on all nodes.
 8.      Create and mount the GPFS file system:
 Example migration to GPFS 2.3 from a previous GPFS version:
 1.      Ensure all disks in the file systems to be migrated are in working order by issuing"mmlsdisk" command and checking for disk status to be "ready" and availability to be "up".
 2.      Shutdown GPFS daemons on all nodes of the cluster
 3.      Export the GPFS file systems (mmexportfs).
 4.      Delete all existing nodes for each nodeset in the cluster (mmdelnode).
 5.      Delete the existing cluster by issuing mmdelcluster. (only a 2.2 or less command)
 6.      Install GPFS 2.3 software on all of the cluster nodes :
 7.      Determine which nodes will be quorum nodes in your GPFS cluster and create a new GPFS cluster across all desired cluster nodes (mmcrcluster).
 8.      Complete the movement of the GPFS file system to your new cluster (mmimportfs).
 9.      Start GPFS on all nodes of the cluster (mmstartup).
 10.  Complete the migration to the new level of GPFS (mmchfs). 
 
 GPFS and Oracle References and Additional Information:
 My Oracle Support
 GPFS
 GPFS 2.3 FAQ:
 GPFS Forum at IBM developerWorks:
 GPFS 2.3 Summary of Changes:
 GPFS 2.3 Announcement Details:http://www-306.ibm.com/common/ssi/OIX.wss? DocURL=http://d03xhttpcl001g.boulder.ibm.com/common/ssi/rep_ca/4/897/ENUS204- 294/index.html&InfoType=AN
 GPFS Product Home Page:
 GPFS 2.3 Publications:
 Direct to man pages for GPFS 2.3 commands:
 Migrating to GPFS 2.3 from previous GPFS versions:
 Establishing disaster recovery for your GPFS cluster:
 EtherChannel and Link Aggregation with AIX 5L
 AIX EtherChannel and IEEE 802.3ad Link Aggregation
 Oracle and AIX 5L
References

My Oracle Support provides customers with access to over a million knowledge articles and a vibrant support community of peers and Oracle experts.