Red Hat ENTERPRISE LINUX 4 - ADMINISTRATION Przewodnik Instalacji

Przeglądaj online lub pobierz Przewodnik Instalacji dla Serwery Red Hat ENTERPRISE LINUX 4 - ADMINISTRATION. Red Hat ENTERPRISE LINUX 4 - ADMINISTRATION Installation guide Instrukcja obsługi

  • Pobierz
  • Dodaj do moich podręczników
  • Drukuj

Podsumowanie treści

Strona 1 - Cluster Administration

LandmannRed Hat Enterprise Linux 4Cluster AdministrationConfiguring and Managing a Red Hat ClusterEdition 1.0

Strona 2 - Edition 1.0

To insert a special character into a gedit file, choose Applications → Accessories →Character Map from the main menu bar. Next, choose Search → Find…

Strona 3

static int kvm_vm_ioctl_deassign_device(struct kvm *kvm, struct kvm_assigned_pci_dev *assigned_dev){ int r = 0; struct

Strona 4 - Table of Contents

Be sure to mention the manual's identifier:Cluster_Administration(EN)-4.8 (2009-5-13T12:45)By mentioning this manual's identifier, we know e

Strona 5

Chapter 1. Red Hat Cluster Configuration and ManagementOverviewRed Hat Cluster allows you to connect a group of computers (called nodes or members) to

Strona 6

Figure 1.1. Red Hat Cluster Hardware Overview1.1.2. Installing Red Hat Cluster softwareTo install Red Hat Cluster software, you must have entitlements

Strona 7

Figure 1.2. Cluster Configuration St ructureThe following cluster configuration tools are available with Red Hat Cluster:Conga — T his is a comprehens

Strona 8

Integration of Cluster Status and LogsFine-Grained Control over User PermissionsThe primary components in Conga are luci and ricci, which are separate

Strona 9 - 1. Document Conventions

Figure 1.3. luci homebase T abFigure 1.4 . luci cluster TabRed Hat Enterprise Linux 4 Cluster Administration14

Strona 10 - 1.2. Pull-quote Conventions

Figure 1.5. luci st orage T ab1.3. system-config-cluster Cluster Administration GUIThis section provides an overview of the cluster administration gra

Strona 11 - 2. Feedback

Figure 1.6. Cluster Configuration ToolThe Cluster Configurat ion T ool represents cluster configuration components in the configuration file(/etc/clus

Strona 12 - Introduction

Red Hat Enterprise Linux 4 Cluster AdministrationConfiguring and Managing a Red Hat ClusterEdition [email protected] m

Strona 13

resources consist of file systems, IP addresses, NFS mounts and exports, and user-createdscripts that are available to any high-availability service i

Strona 14

The nodes and services displayed in the Cluster Status Tool are determined by the clusterconfiguration file (/etc/cluster/cluster.conf). You can use t

Strona 15 - 1.2. Conga

Chapter 2. Before Configuring a Red Hat ClusterThis chapter describes tasks to perform and considerations to make before installing and configuring aR

Strona 16

Table 2.1. Enabled IP Ports on Red Hat Cluster NodesIP PortNumberProtocol Component Reference to Example of iptables Rules6809 UDP cm an (Cluster Mana

Strona 17 - Figure 1.4 . luci cluster Tab

Table 2.2. Enabled IP Ports on a Computer That Runs luciIP PortNumberProtocol Component Reference to Example of iptables Rules8084 TCP luci (Conga use

Strona 18

Example 2.5. Port 16851: modclusterd-A INPUT -i 10.10.10.200 -m state --state NEW -m m ultiport -p tcp -s 10.10.10.0/24 -d 10.10.10.0/24 --dports 168

Strona 19

completely rather than attempting a clean shutdown (for example, shutdown -h now). Otherwise, ifACPI Soft-Off is enabled, an integrated fence device c

Strona 20 - 1.3.2. Cluster Status Tool

NoteThis is the preferred method of disabling ACPI Soft-Off.Disable ACPI Soft-Off with chkconfig management at each cluster node as follows:1. Run ei

Strona 21

NoteYou can fence the node with the fence_node command or Conga.Example 2.11. BIOS CMOS Setup Utility: Soft -Off by PWR-BT T N set to Instant-Off+----

Strona 22 - 2.2. Enabling IP Ports

4. When the cluster is configured and running, verify that the node turns off immediately whenfenced.NoteYou can fence the node with the fence_node c

Strona 23

Legal NoticeCopyright © 2008 Red Hat, Inc.This document is licensed by Red Hat under the Creative Commons Attribution-ShareAlike 3.0 UnportedLicense.

Strona 24

# cd /boo t# mki nitrd -f -v initrd-kernel.i mg kernelFor example, the currently running kernel in the following m kinitrd command is 2.6.9-34.0.2.EL:

Strona 25

failed nodes on its own, and can take much longer to do so than CMAN. The default value forCMAN membership timeout is 10 seconds. Other site-specific

Strona 26 - Important

2.8. General Configuration ConsiderationsYou can configure a Red Hat Cluster in a variety of ways to suit your needs. T ake into account thefollowing

Strona 27

Chapter 3. Configuring Red Hat Cluster With CongaThis chapter describes how to configure Red Hat Cluster software using Conga, and consists of thefoll

Strona 28

NoteTypically, a computer in a server cage or a data center hosts luci; however, a clustercomputer can host luci.4. At the computer running luci, ini

Strona 29 - 2.4. Configuring max_luns

4. Click Subm it. Clicking Submit causes the the Create a new cluster page to be displayedagain, showing the parameters entered in the preceding step

Strona 30

The Post-Fail Delay parameter is the number of seconds the fence daemon (fenced) waitsbefore fencing a node (a member of the fence domain) after the n

Strona 31

NoteClicking Apply on the Quorum Partit ion tab propagates changes to the clusterconfiguration file (/etc/cluster/cluster.conf) in each cluster node.

Strona 32

example, name, IP address, login, and password). Modifying a fence device consists of selecting anexisting fence device and changing parameters for th

Strona 33 - 3.2. Starting luci and ricci

causes the display of menu items for fence device configuration: Add a Fence Device andConfigure a Fence Device.NoteIf this is an initial cluster conf

Strona 34 - 3.3. Creating A Cluster

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

Strona 35

1. At the detailed menu for the cluster (below the clusters menu), click Shared Fence Devices.Clicking Shared Fence Devices causes the display of the

Strona 36

NoteYou can choose from an existing fence device or create a new fence device.5. Click Update m ain fence properties and wait for the change to take

Strona 37

10. Click Update main fence properties and wait for the change to take effect.3.6.3. Deleting a Member from a ClusterTo delete a member from an exist

Strona 38

member at the top of the list is the most preferred, followed by the second member in the list, and soon.NoteChanging a failover domain configuration

Strona 39

4. T o enable setting failover priority of the members in the failover domain, click the Prioritizedcheckbox. With Prioritized checked, you can set t

Strona 40

a node means that the node is a member of the failover domain. If Prioritized is checked, you canadjust the priority in the Priority text box for each

Strona 41

is unchecked. Force Unmount kills all processes using the mount point to free upthe mount when it tries to unmount.Reboot host node if unmount fails

Strona 42

4. Click Subm it. Clicking Submit causes a progress page to be displayed followed by the displayof Resources forcluster name page. That page displays

Strona 43

NoteTo verify the existence of the IP service resource used in a cluster service, you must use the /sbin/ip addr list command on a cluster node. T he

Strona 44

Part it ionsVolume GroupsEach section is set up as an expandable tree, with links to property sheets for specific devices,partitions, and storage enti

Strona 45 - 3.8. Adding Cluster Resources

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

Strona 46

Chapter 4. Managing Red Hat Cluster With CongaThis chapter describes various administrative tasks for managing a Red Hat Cluster and consists of thefo

Strona 47

Make a node leave or join a cluster.Fence a node.Reboot a node.Delete a node.To perform one the functions in the preceding list, follow the steps in t

Strona 48

causes the display of services for the cluster in the center of the page.2. At the right of each service listed on the page, click the Choose a task

Strona 49

Chapter 5. Configuring Red Hat Cluster With system-config-clusterThis chapter describes how to configure Red Hat Cluster software using system -config

Strona 50 - 4.2. Managing Cluster Nodes

Configurat ion T ool on cluster node nano-01, do the following:1. Log in to a cluster node and run system-config-cluster. For example:$ ssh -Y root@

Strona 51

Disk Heurist ic. Table 5.1, “Quorum-Disk Parameters” describes the parameters.ImportantQuorum-disk parameters and heuristics depend on the site enviro

Strona 52

Figure 5.2. Creating A New Configuration4. When you have completed entering the cluster name and other parameters in the NewConfiguration dialog box,

Strona 53 - 5.1. Configuration Tasks

Figure 5.3. T he Cluster Configuration ToolRed Hat Enterprise Linux 4 Cluster Administration54

Strona 54 - # system-config-cluster

Table 5.1. Quorum-Disk ParametersParameter DescriptionUse a Quorum Disk Enables quorum disk. Enables quorum-disk parameters in the NewConfiguration di

Strona 55

httpd_cluster). The cluster alias cannot exceed 15 characters.4. (Optional) T he Config Version value is set to 1 by default and is automatically inc

Strona 56

T 94Table of Contents 3

Strona 57

3. Specify the information in the Fence Device Configuration dialog box according to the typeof fence device. Refer to Appendix B, Fence Device Param

Strona 58

Figure 5.6. Adding a Member to a New GULM Cluster3. At the Cluster Node Name text box, specify a node name. T he entry can be a name or an IPaddress

Strona 59

h. T o create additional fence devices at this fence level, return to step 6d. Otherwise, proceedto the next step.i. T o create additional fence lev

Strona 60

7. Start system -config-cluster (refer to Section 5.2, “Starting the Cluster Configurat ionTool”). At the Cluster Configurat ion T ool tab, verify th

Strona 61

c. service clvm d stop, if CLVM has been used to create clustered volumesd. service fenced stope. service cman stopf. service ccsd stop3. At syst

Strona 62

NoteMake sure to configure other parameters that may be affected by changes in this section. Referto Section 5.1, “Configuration Tasks”.5.5.4. Adding

Strona 63

To delete a member functioning only as a GULM client from an existing cluster that is currently inoperation, follow these steps:1. At one of the runn

Strona 64

d. service gfs start, if you are using Red Hat GFSe. service rgm anager start, if the cluster is running high-availability services(rgm anager)6. A

Strona 65

b. Select the cluster node to be deleted. At the bottom of the right frame (labeled Properties),click the Delete Node button.c. Clicking the Delete

Strona 66

NoteMake sure to configure other parameters that may be affected by changes in this section. Referto Section 5.1, “Configuration Tasks”.5.6. Configuri

Strona 67

Red Hat Enterprise Linux 4 Cluster Administration4

Strona 68

Section 5.6.1, “Adding a Failover Domain”Section 5.6.2, “Removing a Failover Domain”Section 5.6.3, “Removing a Member from a Failover Domain”5.6.1. Ad

Strona 69

Adjusting Priority”). Clicking Prioritized List causes the Priority column to be displayednext to the Member Node column.Figure 5.11. Failover Domain

Strona 70

New cluster — If this is a new cluster, choose File => Save to save the changes to the clusterconfiguration.Running cluster — If this cluster is op

Strona 71

system ID explicitly, specify it in this field.Force Unmount checkbox — If checked, forces the file system to unmount. T he defaultsetting is unchecke

Strona 72 - 5.7. Adding Cluster Resources

Read-Write and Read Only options — Specify the type of access rights for this NFSclient resource:Read-Write — Specifies that the NFS client has read-w

Strona 73

NoteUse a descriptive name that clearly distinguishes the service from other services in thecluster.Figure 5.12. Adding a Cluster Service4. If you wa

Strona 74

7. Select a recovery policy to specify how the resource manager should recover from a servicefailure. At the upper right of the Service Management di

Strona 75

For newly defined clusters, you must propagate the configuration file to the cluster nodes as follows:1. Log in to the node where you created the con

Strona 76

Chapter 6. Managing Red Hat Cluster With system-config-clusterThis chapter describes various administrative tasks for managing a Red Hat Cluster and c

Strona 77

Figure 6.1. Cluster Status ToolYou can use the Cluster Status Tool to enable, disable, restart, or relocate a high-availability service.The Cluster St

Strona 78

IntroductionThis document provides information about installing, configuring and managing Red Hat Clustercomponents. Red Hat Cluster components are pa

Strona 79

Table 6.1. Members Stat usMembers Status DescriptionMemberThe node is part of the cluster.Note: A node can be a member of a cluster; however, the node

Strona 80

ImportantAlthough the Cluster Configuration Tool provides a Quorum Votes parameter in theProperties dialog box of each cluster member, that parameter

Strona 81

file /etc/cluster/cluster.conf.bak.1 is the newest backup, /etc/cluster/cluster.conf.bak.2 is the second newest backup, and /etc/cluster/cluster.conf.

Strona 82

# chkconfig --level 2345 rgmanager o ff# chkconfig --level 2345 gfs off# chkconfig --level 2345 clvmd off# chkconfig --level 2345 fenced off# chkconfi

Strona 83

Example of Setting Up Apache HTTP ServerThis appendix provides an example of setting up a highly available Apache HT T P Server on a Red HatCluster. T

Strona 84 - # chkconfig --del httpd

step. Specify the drive letter and the partition number. For example:# mkfs -t ext3 /dev/sde33. Mount the file system that contains the document root

Strona 85 - Listen 192.168.1.100:80

<Directory /mnt/httpdservice/cgi-bin">AllowOverride NoneOptions None Order allow,deny Allow from all </Directory>Additional changes m

Strona 86

choose each resource that you created in the previous steps. Repeat this step until allresources have been added.Click OK.6. Choose File => Save t

Strona 87 - Click OK

Fence Device ParametersThis appendix provides tables with parameter descriptions of fence devices.NoteCertain fence devices have an optional Password

Strona 88 - Fence Device Parameters

Table B.3. Bull PAP (Platform Administration Processor)Field DescriptionName A name for the Bull PAP system connected to the cluster.IP Address The IP

Strona 89

Red Hat Cluster Suite documentation and other Red Hat documents are available in HTML, PDF, andRPM versions on the Red Hat Enterprise Linux Documentat

Strona 90

Table B.6. Fujitsu Siemens Remoteview Service Board (RSB)Field DescriptionName A name for the RSB to use as a fence device.Hostname The hostname assig

Strona 91

Table B.10. IBM Remote Supervisor Adapter II (RSA II)Field DescriptionName A name for the RSA device connected to the cluster.Hostname The hostname as

Strona 92

WarningManual fencing is not supported for production environments.Table B.14 . McData SAN SwitchField DescriptionName A name for the McData device co

Strona 93

Table B.18. Virt ual Machine FencingField DescriptionName Name of the virtual machine fencing device.Domain Unique domain name of the guest to be fenc

Strona 94 - Revision History

Revision HistoryRevision 1.0- 10.4 00 2013-10-31 Rüdiger LandmannRebuild with publican 4.0.0Revision 1.0- 10 2012-07-18 Anthony T ownsRebuild for Publ

Strona 95

- enabling IP ports, Enabling IP Ports- general considerations, General Configuration Considerations- managing cluster node, Managing Cluster Nodes- m

Strona 96

command line tools table, Command Line Administration Toolsconfiguration file- propagation of, Propagating T he Configuration File: New Clusterconfigu

Strona 97

ipt ables- configuring, Enabling IP PortsMmax_luns- configuring, Configuring max_lunsPparameters, fence device, Fence Device Parameterspower controlle

Komentarze do niniejszej Instrukcji

Brak uwag