Product SiteDocumentation Site

Red Hat Enterprise Linux 5

Configuration Example - NFS Over GFS

Configuring NFS over GFS in a Red Hat Cluster

Edition 3

Logo

Legal Notice

Copyright © 2011 Red Hat Inc..
The text of and illustrations in this document are licensed by Red Hat under a Creative Commons Attribution–Share Alike 3.0 Unported license ("CC-BY-SA"). An explanation of CC-BY-SA is available at http://creativecommons.org/licenses/by-sa/3.0/. In accordance with CC-BY-SA, if you distribute this document or an adaptation of it, you must provide the URL for the original version.
Red Hat, as the licensor of this document, waives the right to enforce, and agrees not to assert, Section 4d of CC-BY-SA to the fullest extent permitted by applicable law.
Red Hat, Red Hat Enterprise Linux, the Shadowman logo, JBoss, MetaMatrix, Fedora, the Infinity Logo, and RHCE are trademarks of Red Hat, Inc., registered in the United States and other countries.
Linux® is the registered trademark of Linus Torvalds in the United States and other countries.
Java® is a registered trademark of Oracle and/or its affiliates.
XFS® is a trademark of Silicon Graphics International Corp. or its subsidiaries in the United States and/or other countries.
MySQL® is a registered trademark of MySQL AB in the United States, the European Union and other countries.
All other trademarks are the property of their respective owners.


1801 Varsity Drive
 Raleigh, NC 27606-2072 USA
 Phone: +1 919 754 3700
 Phone: 888 733 4281
 Fax: +1 919 754 3701

Abstract
This book describes a procedure for configuring NFS over GFS in a 5-node Red Hat Cluster using Conga.
Introduction
1. About This Guide
2. Audience
3. Software Versions
4. Related Documentation
5. Feedback
6. Document Conventions
6.1. Typographic Conventions
6.2. Pull-quote Conventions
6.3. Notes and Warnings
1. NFS over GFS in a Red Hat Cluster
2. Prerequisite Configuration
3. Components to Configure
4. Configuring the Cluster Resources
4.1. Configuring an IP Address Resource
4.2. Configuring a GFS Resource
4.3. Configuring an NFS Export Resource
4.4. Configuring NFS Client Resources
5. Configuring an NFS Cluster Service
5.1. Add a Service to the Cluster
5.2. Adding an IP Address Resource to an NFS Service
5.3. Adding a GFS Resource to an NFS Service
5.4. Adding an NFS Export Resource to an NFS Service
5.5. Adding NFS Client Resources to an NFS Service
6. Testing the NFS Cluster Service
7. Troubleshooting
8. The Cluster Configuration File
9. Configuration Considerations
9.1. Locking Considerations
9.2. Additional Configuration Considerations
A. Revision History
Index

Introduction

1. About This Guide

This book describes a procedure for configuring NFS over GFS in a 5-node Red Hat Cluster using Conga.

2. Audience

This book is intended to be used by system administrators managing systems running the Linux operating system. It requires familiarity with Red Hat Enterprise Linux 5, Red Hat Cluster Suite, GFS file system administration, and basic knowledge of NFS administration.

3. Software Versions

Table 1. Software Versions
Software Description
RHEL5
refers to RHEL5 and higher
GFS
refers to GFS for RHEL5 and higher

5. Feedback

If you spot a typo, or if you have thought of a way to make this manual better, we would love to hear from you. Please submit a report in Bugzilla (http://bugzilla.redhat.com/bugzilla/) against the component Documentation-cluster.
Be sure to mention the manual's identifier: 
Bugzilla component: Documentation-cluster
Book identifier: Cluster_Config_Example_NFS_GFS(EN)-5 (2011-07-21T15:20)
By mentioning this manual's identifier, we know exactly which version of the guide you have.
If you have a suggestion for improving the documentation, try to be as specific as possible. If you have found an error, please include the section number and some of the surrounding text so we can find it easily.

6. Document Conventions

This manual uses several conventions to highlight certain words and phrases and draw attention to specific pieces of information.
In PDF and paper editions, this manual uses typefaces drawn from the Liberation Fonts set. The Liberation Fonts set is also used in HTML editions if the set is installed on your system. If not, alternative but equivalent typefaces are displayed. Note: Red Hat Enterprise Linux 5 and later includes the Liberation Fonts set by default.

6.1. Typographic Conventions

Four typographic conventions are used to call attention to specific words and phrases. These conventions, and the circumstances they apply to, are as follows.
Mono-spaced Bold
Used to highlight system input, including shell commands, file names and paths. Also used to highlight keycaps and key combinations. For example:
To see the contents of the file my_next_bestselling_novel in your current working directory, enter the cat my_next_bestselling_novel command at the shell prompt and press Enter to execute the command.
The above includes a file name, a shell command and a keycap, all presented in mono-spaced bold and all distinguishable thanks to context.
Key combinations can be distinguished from keycaps by the hyphen connecting each part of a key combination. For example:
Press Enter to execute the command.
Press Ctrl+Alt+F2 to switch to the first virtual terminal. Press Ctrl+Alt+F1 to return to your X-Windows session.
The first paragraph highlights the particular keycap to press. The second highlights two key combinations (each a set of three keycaps with each set pressed simultaneously).
If source code is discussed, class names, methods, functions, variable names and returned values mentioned within a paragraph will be presented as above, in mono-spaced bold. For example:
File-related classes include filesystem for file systems, file for files, and dir for directories. Each class has its own associated set of permissions.
Proportional Bold
This denotes words or phrases encountered on a system, including application names; dialog box text; labeled buttons; check-box and radio button labels; menu titles and sub-menu titles. For example:
Choose System → Preferences → Mouse from the main menu bar to launch Mouse Preferences. In the Buttons tab, click the Left-handed mouse check box and click Close to switch the primary mouse button from the left to the right (making the mouse suitable for use in the left hand).
To insert a special character into a gedit file, choose Applications → Accessories → Character Map from the main menu bar. Next, choose Search → Find… from the Character Map menu bar, type the name of the character in the Search field and click Next. The character you sought will be highlighted in the Character Table. Double-click this highlighted character to place it in the Text to copy field and then click the Copy button. Now switch back to your document and choose Edit → Paste from the gedit menu bar.
The above text includes application names; system-wide menu names and items; application-specific menu names; and buttons and text found within a GUI interface, all presented in proportional bold and all distinguishable by context.
Mono-spaced Bold Italic or Proportional Bold Italic
Whether mono-spaced bold or proportional bold, the addition of italics indicates replaceable or variable text. Italics denotes text you do not input literally or displayed text that changes depending on circumstance. For example:
To connect to a remote machine using ssh, type ssh username@domain.name at a shell prompt. If the remote machine is example.com and your username on that machine is john, type ssh john@example.com.
The mount -o remount file-system command remounts the named file system. For example, to remount the /home file system, the command is mount -o remount /home.
To see the version of a currently installed package, use the rpm -q package command. It will return a result as follows: package-version-release.
Note the words in bold italics above — username, domain.name, file-system, package, version and release. Each word is a placeholder, either for text you enter when issuing a command or for text displayed by the system.
Aside from standard usage for presenting the title of a work, italics denotes the first use of a new and important term. For example:
Publican is a DocBook publishing system.

6.2. Pull-quote Conventions

Terminal output and source code listings are set off visually from the surrounding text.
Output sent to a terminal is set in mono-spaced roman and presented thus: 
books        Desktop   documentation  drafts  mss    photos   stuff  svn
books_tests  Desktop1  downloads      images  notes  scripts  svgs
Source-code listings are also set in mono-spaced roman but add syntax highlighting as follows: 
package org.jboss.book.jca.ex1;

import javax.naming.InitialContext;

public class ExClient
{
   public static void main(String args[]) 
       throws Exception
   {
      InitialContext iniCtx = new InitialContext();
      Object         ref    = iniCtx.lookup("EchoBean");
      EchoHome       home   = (EchoHome) ref;
      Echo           echo   = home.create();

      System.out.println("Created Echo");

      System.out.println("Echo.echo('Hello') = " + echo.echo("Hello"));
   }
}

6.3. Notes and Warnings

Finally, we use three visual styles to draw attention to information that might otherwise be overlooked.

Note

Notes are tips, shortcuts or alternative approaches to the task at hand. Ignoring a note should have no negative consequences, but you might miss out on a trick that makes your life easier.

Important

Important boxes detail things that are easily missed: configuration changes that only apply to the current session, or services that need restarting before an update will apply. Ignoring a box labeled 'Important' will not cause data loss but may cause irritation and frustration.

Warning

Warnings should not be ignored. Ignoring warnings will most likely cause data loss.

Chapter 1. NFS over GFS in a Red Hat Cluster

This document provides a procedure to configure an NFS service in a Red Hat cluster using the Conga configuration tool. The configured cluster will have the following characteristics:
  • There are 5 nodes in the cluster.
  • The NFS service runs over a GFS file system.
  • There are five NFS clients.
Note that this configuration is not a "high capacity" configuration in the sense that more than one server is providing NFS service. In this configuration, the floating IP moves about as needed, but only one server is active at a time.
Figure 1.1, “NFS over GFS in a 5-Node Cluster” shows the NFS over GFS configuration that this procedure yields.
NFS over GFS in a 5-Node Cluster
NFS over GFS in a 5-Node Cluster
Figure 1.1. NFS over GFS in a 5-Node Cluster

In this configuration, Cluster Suite assigns which server will run the NFS service. If that server goes down, the NFS service will automatically fail over to another server in the cluster. The client will not be aware of any loss of service.
The NFS resource that this configuration defines will be NFS Version 3 by default. If you need to restrict what NFS protocol your system provides to its clients, you can do this at NFS startup; this is not part of cluster configuration.
This remainder of this document is organized as follows:

Chapter 2. Prerequisite Configuration

This document describes a procedure for configuring cluster resources and a cluster service. Before beginning this procedure, you must configure a cluster with a fencing device configured for each node in the cluster. In addition, you must configure a logical volume with a GFS filesystem that can be shared among the nodes in the cluster.
Table 2.1, “Configuration Prerequisities” summarizes the prequisite configuration components that have been set up before this procedure beings.
Table 2.1. Configuration Prerequisities
Component Name Comment
cluster nfsclust five-node cluster
cluster node clusternode1.example.com node in cluster nfsclust configured with a fencing device of nfs-apc
cluster node clusternode2.example.com node in cluster nfsclust configured with a fencing device of nfs-apc
cluster node clusternode3.example.com node in cluster nfsclust configured with a fencing device of nfs-apc
cluster node clusternode4.example.com node in cluster nfsclust configured with a fencing device of nfs-apc
cluster node clusternode5.example.com node in cluster nfsclust configured with a fencing device of nfs-apc
LVM volume /dev/myvg/myvol The LVM device on which the GFS file system is created
GFS file system The GFS file system to export by means of NFS, built on LVM volume /dev/myvg/myvol, mounted at /mnt/gfs, and shared among the members of cluster nfsclust
IP address 10.15.86.96 The IP address for the NFS service
NFS Client nfsclient1.example.com System that will access the NFS service
NFS Client nfsclient2.example.com System that will access the NFS service
NFS Client nfsclient3.example.com System that will access the NFS service
NFS Client nfsclient4.example.com System that will access the NFS service
NFS Client nfsclient5.example.com System that will access the NFS service

No failover domain has been defined for this cluster. The cluster software determines on which node to run the NFS service.
For information on configuring a cluster, see Configuring and Managing a Red Hat Cluster. For information on configuring an LVM volume, see LVM Administrator's Guide: Configuration and Administration.

Chapter 3. Components to Configure

This procedure documents the configuration of cluster resources and an NFS cluster service that will be named nfssvc.
Table 3.1, “Cluster Resources to Configure” summarizes the cluster resources that this procedure configures.
Table 3.1. Cluster Resources to Configure
Resource Type Resource Name Description
IP Address 10.15.86.96 The IP address for the NFS service
GFS mygfs The GFS file system that will be exported through the NFS service
NFS Export mynfs
NFS Client nfsclient1 The NFS client system nfsclient1.example.com
NFS Client nfsclient2 The NFS client system nfsclient2.example.com
NFS Client nfsclient3 The NFS client system nfsclient3.example.com
NFS Client nfsclient4 The NFS client system nfsclient4.example.com
NFS Client nfsclient5 The NFS client system nfsclient5.example.com

The procedure for configuring these resources and the specific parameters to configure are described in Chapter 4, Configuring the Cluster Resources.
You configure a Cluster service by adding a service and defining the composition of the service, which consists of the resources that the service requires. In this procedure, you will create an NFS cluster service named nfssvc. Table 3.2, “Parameters to Configure for NFS Cluster Service nfssvc” summarizes the resource configuration of nfssvc. The names of the resources are those that you assign when you define them, as noted in Table 3.1, “Cluster Resources to Configure”
Table 3.2. Parameters to Configure for NFS Cluster Service nfssvc
Resource Name Comment
IP Address Resource
GFS Resource mygfs
NFS Export Resource mynfs NFS Export resource mynfs is a child of GFS resource mygfs.
NFS Client Resource nfsclient1 NFS Client resource nfsclient1 is a child of NFS Export resource mynfs.
NFS Client Resource nfsclient2 NFS Client resource nfsclient2 is a child of NFS Export resource mynfs.
NFS Client Resource nfsclient3 NFS Client resource nfsclient3 is a child of NFS Export resource mynfs.
NFS Client Resource nfsclient4 NFS Client resource nfsclient4 is a child of NFS Export resource mynfs.
NFS Client Resource nfsclient5 NFS Client resource nfsclient5 is a child of NFS Export resource mynfs.

The procedure for configuring the NFS service is described in Chapter 5, Configuring an NFS Cluster Service.

Chapter 4. Configuring the Cluster Resources

4.1. Configuring an IP Address Resource
4.2. Configuring a GFS Resource
4.3. Configuring an NFS Export Resource
4.4. Configuring NFS Client Resources
This chapter provides the procedures for configuring the cluster resources needed for an NFS service. For the procedure for configuring the NFS service itself, see Chapter 5, Configuring an NFS Cluster Service.
Before you can configure an NFS service for a GFS file system in a cluster, you must add the following resources to your cluster:
To add a resource to your cluster using Conga, perform the following procedure:
  1. As an administrator of luci Select the cluster tab.
  2. From the Choose a cluster to administer screen, select the cluster to which you will add resources. In this example, that is the cluster with the name nfsclust.
  3. At the menu for cluster nfsclust (below the clusters menu), click Resources. This causes the display of menu items for resource configuration: Add a Resource and Configure a Resource.
  4. Click Add a Resource. Clicking Add a Resource causes the Add a Resource page to be displayed.
The following sections provide instructions for adding the resources you need for an NFS service.

4.1. Configuring an IP Address Resource

Use the following procedure to add the IP Address resource 10.15.86.96 to cluster nfsclust.
  1. At the Add a Resource page for cluster nfsclust, click the drop-down box under Select a Resource Type and select IP Address
  2. For IP Address, enter 10.15.86.96.
  3. Leave the Monitor Link checkbox selected to enable link status monitoring of the IP address resource.
  4. Click Submit. Clicking Submit displays a verification page. Verifying that you want to add this resource displays a progress page followed by the display of Resources page, which displays the resources that have been configured for the cluster.

4.2. Configuring a GFS Resource

Use the following procedure to add the GFS file system resource mygfs to cluster nfsclust.
  1. At the Add a Resource page for cluster nfsclust, click the drop-down box under Select a Resource Type and select GFS file system
  2. For Name, enter mygfs.
  3. For Mount point, enter /mnt/gfs. This is the path to which the GFS file system is mounted.
  4. For Device, enter /dev/myvg/myvol. The is the LVM logical volume on which the GFS file system was is created.
  5. The Options field specifies the mount options for the GFS file system. For this example, we are mounting the file system with the rw (read-write) and localflocks option.
  6. Leave the File System ID field blank. Leaving the field blank causes a file system ID to be assigned automatically after you click Submit at the File System Resource Configuration dialog box.
  7. Leave the Force Unmount checkbox unchecked. Force Unmount kills all processes using the mount point to free up the mount when it tries to unmount. With GFS resources, the mount point is not unmounted at service tear-down unless this box is checked.
  8. Click Submit and accept the verification screen.

4.3. Configuring an NFS Export Resource

Use the following procedure to add NFS export resource mynfs to cluster nfsclust.
  1. At the Add a Resource page for cluster nfsclust, click the drop-down box under Select a Resource Type and select NFS Export
  2. For Name, enter mynfs.
  3. Click Submit and accept the verification screen.
The NFS Eport resource that this configuration defines will be NFS Version 3 by default. If you need to restrict what NFS protocol your system provides to its clients, you can do this at NFS startup; this is not part of cluster configuration.

4.4. Configuring NFS Client Resources

This example procedure configures five NFS client resources for cluster nfsclust. The procedure for configuring the first two clients only is laid out explicitly.
Use the following procedure to add NFS client resource nfsclient1 to cluster nfsclust.
  1. At the Add a Resource page for cluster nfsclust, click the drop-down box under Select a Resource Type and select NFS client
  2. For Name, enter nfsclient1.
  3. For Target, enter nfsclient1.example.com. This is the first NFS client system.
  4. This Options field species additional client access rights. Specify rw (read-write) in this field. For more information, refer to the General Options section of the exports(5) man page.
  5. Check the Allow Recover checkbox. This indicates that if someone removes the export from the export list, the system will recover the export inline without taking down the NFS service.
  6. Click Submit and accept the verification screen.
Use the following procedure to add NFS client resource nfsclient2 to cluster nfsclust.
  1. At the Add a Resource page for cluster nfsclust, click the drop-down box under Select a Resource Type and select NFS client
  2. For Name, enter nfsclient2.
  3. For Target, enter nfsclient2.example.com. This is the second NFS client system.
  4. Leave the Options field blank.
  5. Check the Allow Recover checkbox.
  6. Click Submit and accept the verification screen.
Use the same procedure to configure the remaining three NFS client resources, using nfsclient3, nfsclient4, and nfsclient5 as the names of the resources and using nfsclient3.example.com, nfsclient4.example.com, and nfsclient5.example.com as the targets.

Chapter 5. Configuring an NFS Cluster Service

5.1. Add a Service to the Cluster
5.2. Adding an IP Address Resource to an NFS Service
5.3. Adding a GFS Resource to an NFS Service
5.4. Adding an NFS Export Resource to an NFS Service
5.5. Adding NFS Client Resources to an NFS Service
This chapter provides the procedure for configuring an NFS service in a Red Hat Cluster Suite.
In order to configure an NFS service for a GFS file system in a cluster, you need to perform the following steps:
Some Red Hat Cluster resources must be configured as parent or child resources in a service. A resource that is configured as a parent resource is started before the children of that resource are started. All resources that are configured as a child resources of a parent must stop cleanly before a parent resource may be stopped. For a resource that is configured as a parent resource to be considered in good health, all the resources configured as its child resources must be in good health.
When you configure the NFS export resource for an NFS service, you must configure it as a child of the GFS resource. You must configure each NFS client resource for an NFS service as a child of the NFS export resource.

5.1. Add a Service to the Cluster

To add a service to your cluster using Conga, perform the following procedure:
  1. As an administrator of luci Select the cluster tab.
  2. From the Choose a cluster to administer screen, select the cluster to which you will add resources. In this example, that is the cluster with the name nfsclust.
  3. At the menu for cluster nfsclust (below the clusters menu), click Services. This causes the display of menu items for service configuration: Add a Service and Configure a Service.
  4. Click Add a Service. Clicking Add a Service causes the Add a Service page to be displayed.
  5. For Name, enter nfssvc.
  6. Leave the checkbox labeled Automatically start this service checked, which is the default setting. When the checkbox is checked, the service is started automatically when a cluster is started and running. If the checkbox is not checked, the service must be started manually any time the cluster comes up from the stopped state.
  7. Leave the Run Exclusive checkbox unchecked. The Run Exclusive checkbox sets a policy wherein the service only runs on nodes that have no other services running on them. Since an NFS service consumes few resources, two services could run together on the same node without contention for resources and you do not need to check this.
  8. For Failover Domain, leave the drop-down box default value of None. In this configuration, all of the nodes in the cluster may be used for failover.
  9. For Recovery Policy, the drop-down box displays Select a recovery policy. Click the drop-down box and select relocate. This policy indicates that the system should relocate the service before restarting; it should not restart the node where the service is currently located.
  10. Add the NFS service resources to this resource, as described in the following sections.
  11. After you have added the NFS resources to the service, click Submit. The system prompts you to verify that you want to create this service. Clicking OK causes a progress page to be displayed followed by the display of Services page for the cluster. That page displays the services that have been configured for the cluster.

5.2. Adding an IP Address Resource to an NFS Service

Use the following procedure to add an IP Address resource to the NFS cluster service nfssvc.
  1. At the Add a Service page for cluster nfsclust, click Add a resource to this service. Clicking Add a resource to this service causes the display of two drop-down boxes: Add a new local resource and Use an existing global resource.
    For this example, we will use global resources, which are resources that were previously added as global resources. Adding a new local resource would add a resource that is available only to this service.
  2. In the drop-down box underneath the Use an existing global resource display, click on the Select a resource name display. This displays the resources that have been defined for this cluster.
  3. Select 10.15.86.96 (IP Address). This returns you to the Add a Service page with the IP Address resource displayed.
    Leave the Monitor link checkbox selected, which is the default value. This enables link status monitoring of the IP address resource.

5.3. Adding a GFS Resource to an NFS Service

Use the following procedure to add a GFS resource to the NFS cluster service nfssvc.
  1. At the Add a Service page for cluster nfsclust, click Add a resource to this service.
  2. In the drop-down box underneath the Use an existing global resource display, click on the Select a resource name display.
  3. Select mygfs (GFS). This returns you to the Add a Service page with the GFS resource displayed, with the parameters that you defined in Section 4.2, “Configuring a GFS Resource” displayed.

5.4. Adding an NFS Export Resource to an NFS Service

Configure the NFS Export resource as a child of the GFS resource by following this procedure:
  1. At the Add a Service page for cluster nfsclust, below the GFS Resource Configuration display, click Add a child. This causes the display of two drop-down boxes: Add a new local resource and Use an existing global resource.
  2. In the drop-down box underneath the Use an existing global resource display, click on the Select a resource name display.
  3. Select mynfs (NFS Export). This returns you to the Add a Service page with the NFS Export resource displayed.

5.5. Adding NFS Client Resources to an NFS Service

Configure the NFS Client resources as children of the NFS export resource by following this procedure for each NFS client:
  1. At the Add a Service page for cluster nfsclust, below the NFS Export Resource Configuration display, click Add a child. This causes the display of two drop-down boxes: Add a new local resource and Use an existing global resource.
  2. Click on the Select a resource name display in the drop-down box underneath the Use an existing global resource display.
  3. Select nfsclient1 (NFS Client). This returns you to the Add a Service page with the NFS client resource displayed with the parameters you defined in Section 4.4, “Configuring NFS Client Resources”.
Follow the same procedure to add a second, third, fourth, and fifth NFS client resource, selecting nfsclient2 (NFS Client), nfsclient3 (NFS Client), nfsclient4 (NFS Client), and nfsclient5 (NFS Client) as the resources to add.
After you have added the NFS client resources to the service, you can click Submit. The system prompts you to verify that you want to create this service. Clicking OK causes a progress page to be displayed followed by the display of Services page for the cluster. That page displays the services that have been configured for the cluster.

Chapter 6. Testing the NFS Cluster Service

After you have configured the NFS service, you can check to be sure that the NFS service is working and that it will continue to work as expected if one of the nodes goes down. The following procedure tests an NFS mount on a client, fences the node on which the NFS service is running, and then checks to be sure that the NFS client can still access the file system.
  1. If the GFS file system in the nfsclust cluster is currently empty, populate the file system with test data.
  2. Log in to one of the client systems you defined as a target.
  3. Mount the NFS file system on the client system, and check to see if the data on that file system as available.
  4. On the Luci server, select Nodes from the menu for nfsclust. This displays the nodes in nfsclust and indicates which node is running the nfssvc service.
  5. The drop-down box for each node displays Choose a task. For the node on which the nfssvc service is running, select Fence this node.
  6. Refresh the screen. The nfssvc service should now be running in a different node.
  7. On the client system, check whether the file system you mounted is still available. Even though the NFS service is now running on a different node in the cluster, the client system should detect no difference.
  8. Restore the system to its previous state:
    • Unmount the file system from the client system.
    • Delete any test data you created in the GFS file system.
    • Click on Choose a task in the drop-down box for the node which you fenced and select Reboot this node.

Note

For advanced troubleshooting once a system has gone into production, it is sometimes necessary to set up the netconsole and kdump services on a system. You may find it useful to implement and test these tools before a system goes into production, to help in troubleshooting down the line.

Chapter 7. Troubleshooting

If you find that you are seeing error messages when you try to configure your system, or if after configuration your system does not behave as expected, you can perform the following checks and examine the following areas.
  • Connect to one of the nodes in the cluster and execute the clustat(8) command. This command runs a utility that displays the status of the cluster. It shows membership information, quorum view, and the state of all configured user services.
    The following example shows the output of the clustat(8) command. 
    [root@clusternode4 ~]# clustat
Cluster Status for nfsclust @ Wed Dec  3 12:37:22 2008
Member Status: Quorate

 Member Name                              ID   Status
 ------ ----                              ---- ------
 clusternode5.example.com          1 Online, rgmanager
 clusternode4.example.com          2 Online, Local, rgmanager
 clusternode3.example.com          3 Online, rgmanager
 clusternode2.example.com          4 Online, rgmanager
 clusternode1.example.com          5 Online, rgmanager

 Service Name             Owner (Last)                     State
 ------- ---              ----- ------                     -----
 service:nfssvc           clusternode2.example.com         starting
    In this example, clusternode4 is the local node since it is the host from which the command was run. If rgmanager did not appear in the Status category, it could indicate that cluster services are not running on the node.
  • Connect to one of the nodes in the cluster and execute the group_tool(8) command. This command provides information that you may find helpful in debugging your system. The following example shows the output of the group_tool(8) command. 
    [root@clusternode1 ~]# group_tool
type             level name       id       state
fence            0     default    00010005 none
[1 2 3 4 5]
dlm              1     clvmd      00020005 none
[1 2 3 4 5]
dlm              1     rgmanager  00030005 none
[3 4 5]
dlm              1     mygfs      007f0005 none
[5]
gfs              2     mygfs      007e0005 none
[5]
    The state of the group should be none. The numbers in the brackets are the node ID numbers of the cluster nodes in the group. The clustat shows which node IDs are associated with which nodes. If you do not see a node number in the group, it is not a member of that group. For example, if a node ID is not in dlm/rgmanager group, it is not using the rgmanager dlm lock space (and probably is not running rgmanager).
    The level of a group indicates the recovery ordering. 0 is recovered first, 1 is recovered second, and so forth.
  • Connect to one of the nodes in the cluster and execute the cman_tool nodes -f command This command provides information about the cluster nodes that you may want to look at. The following example shows the output of the cman_tool nodes -f command. 
    [root@clusternode1 ~]# cman_tool nodes -f
Node  Sts   Inc   Joined               Name
   1   M    752   2008-10-27 11:17:15  clusternode5.example.com
   2   M    752   2008-10-27 11:17:15  clusternode4.example.com
   3   M    760   2008-12-03 11:28:44  clusternode3.example.com
   4   M    756   2008-12-03 11:28:26  clusternode2.example.com
   5   M    744   2008-10-27 11:17:15  clusternode1.example.com
    The Sts heading indicates the status of a node. A status of M indicates the node is a member of the cluster. A status of X indicates that the node is dead. The Inc heading indicating the incarnation number of a node, which is for debugging purposes only.
  • Check whether the cluster.conf is identical in each node of the cluster. If you configure your system with Conga, as in the example provided in this document, these files should be identical, but one of the files may have accidentally been deleted or altered.
  • In addition to using Conga to fence a node in order to test whether failover is working properly as described in Chapter 6, Testing the NFS Cluster Service, you could disconnect the ethernet connection between cluster members. You might try disconnecting one, two, or three nodes, for example. This could help isolate where the problem is.
  • If you are having trouble mounting or modifying an NFS volume, check whether the cause is one of the following:
    • The network between server and client is down.
    • The storage devices are not connected to the system.
    • More than half of the nodes in the cluster have crashed, rendering the cluster inquorate. This stops the cluster.
    • The GFS file system is not mounted on the cluster nodes.
    • The GFS file system is not writable.
    • The IP address you defined in the cluster.conf is not bounded to the correct interface / NIC (sometimes the ip.sh script does not perform as expected).
  • Execute a showmount -e command on the node running the cluster service. If it shows up the right 5 exports, check your firewall configuration for all necessary ports for using NFS.
  • If SELinux is currently in enforcing mode on your system, check your /var/log/audit.log file for any relevant messages. If you are using NFS to serve home directories, check whether the correct SELinux boolean value for nfs_home_dirs has been set to 1; this is required if you want to use NFS-based home directories on a client that is running SELinux. If you do not set this value on, you can mount the directories as root but cannot use them as home directories for your users.
  • Check the /var/log/messages file for error messages from the NFS daemon.
  • If you see the expected results locally at the cluster nodes and between the cluster nodes but not at the defined clients, check the firewall configuration at the clients.

Chapter 8. The Cluster Configuration File

Configuring a cluster with Conga modifies the cluster configuration file. This chapter shows the cluster configuration file before and after the procedures documented in Chapter 4, Configuring the Cluster Resources and Chapter 5, Configuring an NFS Cluster Service were performed.
Before the cluster resources and service were configured, the cluster.conf file appeared as follows. 

<?xml version="1.0"?>
<cluster alias="nfsclust" config_version="1" name="nfsclust">
        <fence_daemon post_fail_delay="0" post_join_delay="3"/>
        <clusternodes>
                <clusternode name="clusternode1.example.com" nodeid="1" votes="1">
                        <fence>
                                <method name="apc-nfs">
                                        <device name="apc1" switch="3" port="1"/>
                                </method>
                        </fence>
                </clusternode>

                <clusternode name="clusternode2.example.com" nodeid="2" votes="1">
                        <fence>
                                <method name="apc-nfs">
                                        <device name="apc1" switch="3" port="2"/>
                                </method>
                        </fence>
                </clusternode>
                <clusternode name="clusternode3.example.com" nodeid="3" votes="1">
                        <fence>
                                <method name="apc-nfs">
                                        <device name="apc1" switch="3" port="3"/>
                                </method>
                        </fence>
                </clusternode>
                <clusternode name="clusternode4.example.com" nodeid="4" votes="1">
                        <fence>
                                <method name="apc-nfs">
                                        <device name="apc1" switch="3" port="4"/>
                                </method>
                        </fence>
                </clusternode>
                <clusternode name="clusternode5.example.com" nodeid="5" votes="1">
                        <fence>
                                <method name="apc-nfs">
                                        <device name="apc1" switch="3" port="5"/>
                                </method>
                        </fence>
                </clusternode>
        </clusternodes>
        <cman/>
        <fencedevices>
                <fencedevice name="apc1" agent="fence_apc" ipaddr="link-apc" login="apc"
                   passwd="apc"/>
        </fencedevices>
        <rm/>
</cluster>
After the cluster resources and service were configured, the cluster.conf file appeared as follows. 

<?xml version="1.0"?>
<cluster alias="nfsclust" config_version="10" name="nfsclust">
        <fence_daemon clean_start="0" post_fail_delay="0" post_join_delay="3"/>
        <clusternodes>
                <clusternode name="clusternode1.example.com" nodeid="1" votes="1">
                        <fence>
                                <method name="apc-nfs">
                                        <device name="apc1" switch="3" port="1"/>
                                </method>
                        </fence>
                </clusternode>
                <clusternode name="clusternode2.example.com" nodeid="2" votes="1">
                        <fence>
                                <method name="apc-nfs">
                                        <device name="apc1" switch="3" port="2"/>
                                </method>
                        </fence>
                </clusternode>
                <clusternode name="clusternode3.example.com" nodeid="3" votes="1">
                        <fence>
                                <method name="apc-nfs">
                                        <device name="apc1" switch="3" port="3"/>
                                </method>
                        </fence>
                </clusternode>
                <clusternode name="clusternode4.example.com" nodeid="4" votes="1">
                        <fence>
                                <method name="apc-nfs">
                                        <device name="apc1" switch="3" port="4"/>
                                </method>
                        </fence>
                </clusternode>
                <clusternode name="clusternode5.example.com" nodeid="5" votes="1">
                        <fence>
                                <method name="apc-nfs">
                                        <device name="apc1" switch="3" port="5"/>
                                </method>
                        </fence>
                </clusternode>
        </clusternodes>
        <cman/>
        <fencedevices>
                <fencedevice name="apc1" agent="fence_apc" ipaddr="link-apc" login="apc"
                   passwd="apc"/>
        </fencedevices>
        <rm>
                <failoverdomains/>
                <resources>
                        <ip address="10.15.86.96" monitor_link="1"/>
                        <clusterfs device="/dev/myvg/myvol" force_unmount="0"
                           fsid="39669" fstype="gfs" mountpoint="/mnt/gfs" 
                           name="mygfs" options="rw,localflocks"/>
                        <nfsexport name="mynfs"/>
			<nfsclient allow_recover="1" name="nfsclient1" options="rw"
                           target="nfsclient1.example.com"/>
                        <nfsclient allow_recover="1" name="nfsclient2" options="rw"
                           target="nfsclient2.example.com"/>
                        <nfsclient allow_recover="1" name="nfsclient3" options="rw"
                           target="nfsclient3.example.com"/>
                        <nfsclient allow_recover="1" name="nfsclient4" options="rw"
                           target="nfsclient4.example.com"/>
                        <nfsclient allow_recover="1" name="nfsclient5" options="rw"
                           target="nfsclient5.example.com"/>
                </resources>
                <service autostart="1" exclusive="0" name="nfssvc" recovery="relocate">
                        <ip ref="10.15.86.96"/>
                        <clusterfs ref="mygfs">
                                <nfsexport ref="mynfs">
                                        <nfsclient ref="nfsclient1"/>
                                        <nfsclient ref="nfsclient2"/>
                                        <nfsclient ref="nfsclient3"/>
                                        <nfsclient ref="nfsclient4"/>
                                        <nfsclient ref="nfsclient5"/>
                                </nfsexport>
                        </clusterfs>
                </service>
        </rm>
</cluster>

Chapter 9. Configuration Considerations

9.1. Locking Considerations
9.2. Additional Configuration Considerations
This chapter describes the caveats you should take into account when configuring an NFS service over a GFS or GFS2 file system.

9.1. Locking Considerations

Warning

When a GFS or GFS2 filesystem is exported via NFS then you must mount the filesystem with the localflocks option. The intended effect of this is to allow the NFS server to manage locks on the GFS or GFS2 filesystem without the extra overhead of passing through the GFS and GFS2 locking layers.
For more information on the localflocks mount option and when it may be required, see the Global File System and Global File System 2 manuals.

9.2. Additional Configuration Considerations

In addition to the locking considerations, you should take the following into account when configuring an NFS service over a GFS or GFS2 file system.
  • Red Hat supports only Red Hat Cluster Suite configurations using NFSv3 with locking in an active/passive configuration with the following characteristics:
    • The backend file system is a GFS or GFS2 file system running on a 2 to 16 node cluster.
    • An NFSv3 server is defined as a service exporting the entire GFS/GFS2 file system from a single cluster node at a time.
    • The NFS server can fail over from one cluster node to another (active/passive configuration).
    • No access to the GFS/GFS2 file system is allowed except through the NFS server. This includes both local GFS/GFS2 file system access as well as access through Samba or Clustered Samba.
    • The GFS or GFS2 file system must be mounted with the localflocks option.
    • There is no NFS quota support on the system.
    This configuration provides HA for the file system and reduces system downtime since a failed node does not result in the requirement to execute the fsck command when failing the NFS server from one node to another.
  • The fsid= NFS option is mandatory for NFS exports of GFS/GFS2.
  • There is currently an issue with failover and failback when using NFSv3 over GFS with TCP when the following scenario comes into play:
    1. Client A mounts from server 1.
    2. The system administrator moves NFS service from server 1 to server 2.
    3. The client resumes I/O operations.
    4. The system administrator moves NFS service from server 2 to server 1.
    In this situation, the NFS service on server 1 does not get shut down because this would render other NFS services inoperable.
    Should this situation arise, you should move all NFS services off of server 1 and run the service nfs restart. After this you can safely migrate your NFS services back to server 1.
  • If problems arise with your cluster (for example, the cluster becomes inquorate and fencing is not successful), the clustered logical volumes and the GFS/GFS2 file system will be frozen and no access is possible until the cluster is quorate. You should consider this possibility when determining whether a simple failover solution such as the one defined in this procedure is the most appropriate for your system.

Revision History

Revision History
Revision 1.0-2Thu Jul 21 2011Steven Levine
Resolves: #676406
Clarifies information about necessity for mounting GFS file systems with the localflocks option.
Revision 1.0-1Tue Aug 3 2010Steven Levine
Resolves: #562251
Clarifies support issues in "Configuration Considerations" chapter.
Resolves: #614526
Adds note about issues of Samba support to "Configuration Considerations" chapter.
Revision 1.0Thu Jan 29 2009Steven Levine
First edition

Index

A

Allow Recover checkbox, Configuring NFS Client Resources

C

child resource
configuration, Adding an NFS Export Resource to an NFS Service, Adding NFS Client Resources to an NFS Service
definition, Configuring an NFS Cluster Service
clustat command, Troubleshooting
cluster nodes, prerequisite, Prerequisite Configuration
cluster resources, Components to Configure
cluster service
adding GFS resource, Adding a GFS Resource to an NFS Service
adding IP address resource, Adding an IP Address Resource to an NFS Service
adding NFS client resource, Adding NFS Client Resources to an NFS Service
adding NFS export resource, Adding an NFS Export Resource to an NFS Service
adding to cluster, Add a Service to the Cluster
composition, Components to Configure
cluster, prerequisite, Prerequisite Configuration
cluster.conf file, The Cluster Configuration File
cman_tool command, Troubleshooting

F

failover domain, Prerequisite Configuration, Add a Service to the Cluster
feedback, Feedback
Force Unmount checkbox, Configuring a GFS Resource

G

getenforce command, Troubleshooting
GFS file system
adding resource to cluster, Configuring a GFS Resource
options, Configuring a GFS Resource
prerequisite, Prerequisite Configuration
resource, Components to Configure
GFS resource
adding to cluster service, Adding a GFS Resource to an NFS Service
group_tool command, Troubleshooting

I

IP address
adding resource to cluster, Configuring an IP Address Resource
prerequisite, Prerequisite Configuration
resource, Components to Configure
IP address resource
adding to cluster service, Adding an IP Address Resource to an NFS Service
IP, floating, NFS over GFS in a Red Hat Cluster

K

kdump service, Testing the NFS Cluster Service

L

LVM volume, prerequisite, Prerequisite Configuration

M

Monitor Link checkbox, Configuring an IP Address Resource, Adding an IP Address Resource to an NFS Service

N

netconsole service, Testing the NFS Cluster Service
NFS
failover and failback issues, Locking Considerations
in a cluster, Locking Considerations
locks, Locking Considerations
startup, NFS over GFS in a Red Hat Cluster
Version 3, NFS over GFS in a Red Hat Cluster
NFS client
adding resource to cluster, Configuring NFS Client Resources
options, Configuring NFS Client Resources
resource, Components to Configure
NFS client resource
adding to cluster service, Adding NFS Client Resources to an NFS Service
NFS client systems, prerequisite, Prerequisite Configuration
NFS export
adding resource to cluster, Configuring an NFS Export Resource
resource, Components to Configure
NFS export resource
adding to cluster service, Adding an NFS Export Resource to an NFS Service
NFS service
configuration overview, Configuring an NFS Cluster Service
testing, Testing the NFS Cluster Service

P

parent resource
definition, Configuring an NFS Cluster Service
POSIX locks, Locking Considerations
Prerequisite Configuration Components, Prerequisite Configuration

R

recovery policy, Add a Service to the Cluster
resource
adding to cluster, Configuring the Cluster Resources
GFS file system, Configuring a GFS Resource
IP address, Configuring an IP Address Resource
NFS client, Configuring NFS Client Resources
NFS export, Configuring an NFS Export Resource

S

SELinux enforcing mode, Troubleshooting
SELinux nfs_home_dirs value, Troubleshooting
service
adding to cluster, Add a Service to the Cluster
automatic start, Add a Service to the Cluster
showmount command, Troubleshooting

T

testing
NFS service, Testing the NFS Cluster Service