GPFs

Run this command on all disks to be used for the GPFS file system.

The new system also benefits from a file system known as GPFS that was developed at IBM Almaden to enable supercomputers faster data access.

It allows users to perform SNMP service collection from a defined Linux OS collector node in the GPFS cluster to a management node.

The combination of GPFS and the DS4500 provides for a redundant storage subsystem in case of disk failure.

GPFS allows both the master and backup nodes to have concurrent access to all of the files in the cluster.

With GPFS, for the file system to remain available, a quorum of nodes needs to be available.

These include the installation of the GPFS code on each node and the configuration of the Qlogic adapters for the storage nodes.

The GPFS portability layer (PL) is kernel-specific, and it must be created separately for each operating system level within your cluster.

GPFS uses this internode access to allow any node in the GPFS cluster to run relevant commands on other nodes.

For this example, two main CSM groups are defined for use during the GPFS configuration, as shown below.

This file system is shared out to the other nodes, through GPFS or NFS.

It includes one large GPFS cluster split into two logical halves with a single large file system.

GPFS also enables a large system to keep track of its many files without laboriously scanning through every one.

After this, with automount enabled, it mounts automatically when a node starts GPFS.

This section covers in detail the steps taken in the creation of a GPFS cluster.

The example setup uses CSM to install MSJ on every storage node as part of the GPFS installation.

For the example, one large GPFS file system is created using all the NSDs defined to GPFS.

It allows users shared file access within a single GPFS cluster and across multiple GPFS clusters.

GPFS requires each cluster member to have a base set of GPFS RPMs installed.

The pagepool mechanism allows GPFS to implement read, as well as write, requests asynchronously.

You create the GPFS cluster for this example using several distinct steps.

Pick a GPFS cluster node to be the collector node.

Instructions on installing and configuring GPFS.

Listing 1 shows output for a typical trap capture with snmptrapd; Listing 2 shows typical GPFS cluster information gathered from snmpwalk.

A single network, a LAN or a switch, is used for GPFS communication, including the NSD communication.

Figure 1 shows the GPFS cluster configuration results used in this article.

When you created the GPFS file system, you should have specified the number of copies of data and metadata as two.

The -r sets the full path for the remote shell program to be used by GPFS.

There are four NSD nodes (also known here as storage nodes) in total: two in each logical half of the GPFS cluster.

The GPFS NSD component provides a method for cluster-wide disk naming and access.

A shared network for GPFS communications allowing a single network view of the configuration.

On the collector node, verify that the GPFS SNMP sub-agent is running: ps -ef | grep mmsnmpagentd.

If the MIB definition file changes in the GPFS build images, repeat this step.

This example has GPFS replicate the data and metadata on the GPFS file system.

The superuser can display the status of the quotas for the file system using the command mmrepquota gpfs.

(See the Distributing the GPFS portability layer section for more details).

This means that most of the members of the GPFS cluster access the storage over IP using the NSD servers.

A list of GPFS prerequisites for this exercise.

Let's also look at a few native commands related to GPFS cluster and TSA PD.

GPFS requires that all nodes in the GPFS cluster have the ability to access each other using the root ID with no password provided.

Storage server cache settings can impact GPFS performance if they are not set correctly.

All disks used by GPFS must first be given a globally-accessible NSD name.

However, if the value is set too high, GPFS will not start.

Instructions on creating the GPFS file system.

GPFS required packages is copied on each instance.

The -R sets the remote file copy program to be used by GPFS.

Start GPFS on all compute nodes from the management node.

See the GPFS network tuning section in Part 4 of the series for more details on the storage network.

GPFS is the distributed file system running on the DS4500 storage server.

PATH has been changed to include the GPFS binary directory.

Both PCMs provide support of POWERHA and GPFS clustered software products.

Listing 1 shows an example script that can be used to configure GPFS.

The level of GPFS used for the example installation was 2. 3. 0-3.

Versions of GPFS before Version 2. 3 do require those manual steps.

WPAR Manager now supports backing up of files in writable, namefs-mounted non-AIX filesystems such as GPFS.

Start GPFS on all of the NSD servers at the same time to prevent NSDs from being marked as down.

The next step is to start GPFS cluster-wide following these steps

Read this section of the article in parallel with the GPFS documentation (see Resources), in particular the following

Network shared disks (NSDs) created and maintained by the NSD component of GPFS

List the GPFS FS name and mount point on which the FS is mounted: db2cluster

GPFS reduces your total cost of ownership

GPFS General Purpose Forecasting System

On AIX machines running GPFS, you can give an NSD name to

Query the collector node for GPFS SNMP info

Find the gpfs and tsamp filesets installed, and find if iocp is available: lslpp