Red Hat Enterprise Linux 5Cluster AdministrationConfiguring and Managing a Red Hat ClusterEdition 5
To connect to a remote machine using ssh, type ssh [email protected] at ashell prompt. If the remote machine is example.com and your username on th
IP Address The IP address assigned to the device.Login The login name used to access the device.Password The password used to authenticate the connect
Port The switch outlet number.fence_wtiThe fence agent for the WTI Network Power Switch.Field Descrip t io nT ab le B.23. SCSI Fen cin gField Descrip
Use SSLconnectionsUse SSL connections to communicate with the device.Power wait Number of seconds to wait after issuing a power off or power on comman
HA Resource ParametersThis appendix provides descriptions of HA resource parameters. You can configure the parameterswith Luci, system-config-cluster,
T ab le C.2. Ap ach e ServerField Descrip t io nName The name of the Apache Service.Server Root The default value is /etc/httpd.Config File Specifies
T ab le C.4 . G FSField Descrip t io nName The name of the file system resource.Mount point The path to which the file system resource is mounted.Devi
Fence thenode if it isunable toclean up LVMtagsFence the node if it is unable to clean up LVM tags.Field Descrip t io nT ab le C.7. MySQ LField Descri
Name Symbolic name for the NFS mount.NoteThis resource is required only when a cluster service is configured to be anNFS client.Mount point Path to wh
T ab le C.13. O racle D BField Descrip t io nInstance name(SID) of OracleinstanceInstance name.Oracle usernameThis is the user name of the Oracle user
Database type Specifies one of the following database types: Oracle, DB6, or ADA.Oracle TNSlistener nameSpecifies Oracle TNS listener name.ABAP stack
Finally, we use three visual styles to draw attention to information that might otherwise be overlooked.NoteNotes are tips, shortcuts or alternative a
Number ofseconds towait beforecheckingstartup statusSpecifies the number of seconds to wait before checking the startup status (donot wait for J2EE-Ad
T ab le C.20. ServiceField Descrip t io nService name Name of service. This defines a collection of resources, known as a resourcegroup or cluster ser
Field Descrip t io nName Specifies a service name for logging and other purposes.Config File Specifies the absolute path to the configuration file. Th
VM MigrationMappingSpecifies an alternate interface for migration. You can specify this when, forexample, the network address used for virtual machine
HA Resource BehaviorThis appendix describes common behavior of HA resources. It is meant to provide ancillaryinformation that may be helpful in config
A cluster service is an integrated entity that runs under the control of rgmanager. All resources in aservice run on the same node. From the perspecti
For more information about typed child resource start and stop ordering, refer to Section D.2.1,“ Typed Child Resource Start and Stop Ordering” . For
<lvm name="1" .../> <ip address="10.1.1.1" .../> <fs name="1" .../> <lvm name="2" ..
D.2.2. Non-t yped Child Resource St art and St op OrderingAdditional considerations are required for non-typed child resources. For a non-typed child
started after nontypedresource:foo. (Non-typed resources are started in the order thatthey appear in the Service resource.)No n-t ype d Child Reso urc
Chapter 1. Red Hat Cluster Configuration and ManagementOverviewRed Hat Cluster allows you to connect a group of computers (called nodes or members) to
<service name="foo"> <fs name="1" mountpoint="/mnt/foo" device="/dev/sdb1" fsid="12344">
Example D.6 . Service foo No rmal Failu re Reco very<service name="foo"> <script name="script_one" ...>
Act io n Syn t axDisplay theresourcerules that rg_testunderstands.rg_test rulesTest aconfiguration (and/usr/share/cluster) forerrors orredundantresour
Cluster Service Resource Check and Failover TimeoutThis appendix describes how rgmanager monitors the status of cluster resources, and how tomodify th
indeterminately long amount of time to start or stop. Unfortunately, a failure to stop (including atimeout) renders the service inoperable (failed sta
NoteFor two node auto-detection in cman, the number of physical nodes matters and not thepresence of the two_node=1 directive in cluster.conf.Clust er
High Availabilty LVM (HA-LVM)The Red Hat High Availability Add-On provides support for high availability LVM volumes (HA-LVM)in a failover configurati
described in Section F.2, “ Configuring HA-LVM Failover with Tagging” .F.1. Configuring HA-LVM Failover wit h CLVM (preferred, Red HatEnt erprise Linu
<lvm ref="lvm"/> <fs ref="FS"/> </service></rm>F.2. Configuring HA-LVM Failover wit h T agg
NoteIf there are multiple logical volumes in the volume group, then the logical volume name(lv_name) in the lvm resource should be left blank or unspe
Fibre Channel switch — A Fibre Channel switch provides access to Fibre Channel storage. Otheroptions are available for storage according to the type o
Upgrading A Red Hat Cluster from RHEL 4 to RHEL 5This appendix provides a procedure for upgrading a Red Hat cluster from RHEL 4 to RHEL 5. Theprocedur
a. At a cluster node, open /etc/cluster/cluster.conf with a text editor.b. If your cluster is configured with GULM as the cluster manager, remove th
Example G .1. G U LM XML Elemen t s an d Co n t en t<gulm> <lockserver name="gulmserver1"/> <lockserver name="gulmser
Revision HistoryRevisio n 10.0- 8 Mo n Sep 8 2014 St even Levin eVersion for 5.11 GA releaseRevisio n 10.0- 5 Mo n Ju n 30 2014 St even Levin
Resolves: #742312Documents support for IBM iPD U Fence DeviceResolves: #757902Documents backup and restore of luci configuration using luci_admin comm
Resolves: #571557Adds note on managing virtual machines in a cluster.Resolves: #742310Documents new privilege level parameter for IPMI fence device.Re
Resolves: #713256Documents new fence_vmware_soap agent.Resolves: #446137Documents procedure to configure a system to listen to luci from the internal
Revisio n 3.0- 1 T u e Au g 18 2009 Pau l K en n ed yResolves: #516128Adds notes about not supporting IPV6.Resolves: #482936Corrects Section 5.7 tit
Cisco MDS, Fen ce Device Paramet ersCisco UC S, Fen ce D evice Paramet ersclu st er- administration, Before Configuring a Red Hat Cluster, Managing
clu st er reso u rce st at u s ch eck, Clu st er Service R eso u rce Ch eck an d Failo verT imeo u tclu st er reso u rce t yp es, Co n sid erat io n
1. Shut down all cluster services on a single cluster node. For instructions on stopping clustersoftware on a node, refer to Section 6.1, “ Starting
f eed b ack, Feed b ackf en ce d evice- APC power switch, Fence Device Parameters- APC power switch over SNMP, Fence Device Parameters- Brocade fabric
IIB M Blad e C en t er, Fen ce D evice Paramet ersIB M iPDU, Fen ce D evice Paramet ersIB M Remo t e Su p erviso r Adap t er II ( RSA II) , Fen ce D e
O racle list en er reso u rce ag en t , HA Reso u rce Paramet ersPp aramet ers, f en ce d evice, Fen ce Device Paramet ersp aramet ers, H A reso u rce
st art in g t h e clu st er so f t ware, St art in g t h e Clu st er So f t warest at u s ch eck, clu st er reso u rce, Clu st er Service Reso u rce
The following cluster configuration tools are available with Red Hat Cluster:Co n g a — This is a comprehensive user interface for installing, configu
To administer a cluster or storage, an administrator adds (or registers) a cluster or a computer to alu ci server. When a cluster or a computer is reg
Fig u re 1.4 . lu ci clu st er T abChapt er 1 . Red Hat Clust er Co nfigurat ion and Management Overview13
Fig u re 1.5. lu ci st o rag e T ab1.3. system-config-cluster Clust er Administ rat ion GUIThis section provides an overview of the cluster administra
NoteWhile system-config-cluster provides several convenient tools for configuring andmanaging a Red Hat Cluster, the newer, more comprehensive tool, C
The Clu st er Co n f ig u rat io n T o o l represents cluster configuration components in theconfiguration file (/etc/cluster/cluster.conf) with a hi
Fig u re 1.7. Clu st er St at u s T o o lThe nodes and services displayed in the Clu st er St at u s T o o l are determined by the clusterconfiguratio
T ab le 1.1. Co mman d Lin e T o o lsCo mman d Lin eT o o lUsed Wit h Pu rp o seccs_tool —ClusterConfigurationSystem ToolClusterInfrastructureccs_t
Chapter 2. Before Configuring a Red Hat ClusterThis chapter describes tasks to perform and considerations to make before installing andconfiguring a R
configurations in which some nodes have access to the file system and others do not arenot supported.This does not require that all nodes actually mou
Section 2.3.1, “Enabling IP Ports on Cluster Nodes”Section 2.3.2, “ Enabling IP Ports on Computers That Run lu ci”2.3.1. Enabling IP Port s on Clust e
T ab le 2.2. En ab led IP Po rt s o n a C o mp u t er T h at Ru n s lu ciIP Po rt N u mb er Pro t o co l Co mpo n en t8084 TCP lu ci (Co n g a use
NoteDisabling ACPI Soft-Off with the BIOS may not be possible with some computers.Appending acpi=off to the kernel boot command line of the /boot/grub
NoteYou can fence the node with the fence_node command or Co n g a.2.4 .2. Disabling ACPI Soft -Off with t he BIOSThe preferred method of disabling AC
| Suspend Mode [Disabled] | || HDD Power Down [Disabled] | || Soft-Off by
Red Hat Enterprise Linux 5 Cluster AdministrationConfiguring and Managing a Red Hat ClusterEdition 5
# grub.conf generated by anaconda## Note that you do not have to rerun grub after making changes to this file# NOTICE: You have a /boot partition. T
Fig u re 2.1. Web Server Clu st er Service Examp leClients access the HA service through the IP address 10.10.10.201, enabling interaction with the w
An HA service is a group of cluster resources configured into a coherent entity that providesspecialized services to clients. An HA service is represe
Parent, child, and sibling relationships among resourcesThe types of resources and the hierarchy of resources depend on the type of service you arecon
CMAN memb ersh ip t imeo u t valu eThe CMAN membership timeout value (the time a node needs to be unresponsive beforeCMAN considers that node to be
NoteWhen using SELinux with Red Hat Cluster Suite in a VM environment, you should ensure thatthe SELinux boolean fenced_can_network_connect is persist
For gnbd:iptables -I INPUT -p tcp -m state --state NEW -m multiport --dports 14567 -j ACCEPTFor luci:iptables -I INPUT -p tcp -m state --state NEW -m
The non-default location for virtual machine configuration files may be anywhere. The advantage ofusing an NFS share or a shared GFS or GFS2 file syst
Chapter 3. Configuring Red Hat Cluster With CongaThis chapter describes how to configure Red Hat Cluster software using C o n g a, and consists of the
# service ricci startStarting ricci: [ OK ]3. Select a computer to host lu ci and install the lu ci soft
Legal NoticeCopyright © 20 14 Red Hat Inc..This do cument is licensed by Red Hat under the Creative Co mmons Attributio n-ShareAlike 3.0Unpo rted Li
1. As administrator of lu ci, select the clu st er tab.2. Click Creat e a New Clu st er.3. At the Clu st er Name text box, enter a cluster name. Th
The Po st - Fail D elay parameter is the number of seconds the fence daemon (fenced)waits before fencing a node (a member of the fence domain) after t
4. Q u o ru m Part it io n tab — This tab provides an interface for configuring these Q u o ru mPart it io n Co n f ig u rat io n parameters: Do
Heu rist icsPat h t o Pro g ram — The program used to determine if this heuristic isalive. This can be anything that can be executed by /bin/sh -c.
1. At the detailed menu for the cluster (below the clu st ers menu), click Sh ared Fen ceDevices. Clicking Sh ared Fen ce Devices causes the displa
4. Specify the information in the Fencing Type dialog box according to the type of fencedevice. Refer to Appendix B, Fence Device Parameters for more
Creating a cluster consists of selecting a set of nodes (or members) to be part of the cluster. Onceyou have completed the initial step of creating a
A progress page shows the progress of those actions for each added node.5. When the process of adding a node is complete, a page is displayed providi
b. On that page, at the Ch o o se a t askdrop-down box, choose to either disable theservice are start it on another node and click Go.c. Upon confir
NoteFailover domains are not required for operation.By default, failover domains are unrestricted and unordered.In a cluster with several members, usi
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
5. To restrict failover to members in this failover domain, click the checkbox next to Rest rictf ailo ver t o t h is d o main ' s members. Wit
6. Restricted failover — To enable or disable restricted failover for members in this failoverdomain, click the checkbox next to R est rict f ailo v
1. At the detailed menu for the cluster (below the clu st ers menu), click Services. ClickingServices causes the display of services in the center of
5. Add a resource to the service; click Add a resource to this service. Clicking Add aresource to this service causes the display of two drop-down bo
Configuring a Virtual Machine ServiceTo configure a virtual machine service, after clicking Services you can click Ad d a Virt u alMach in e Service.
NoteShared storage for use in Red Hat Cluster Suite requires that you be running the clusterlogical volume manager daemon (clvmd) or the High Availabi
Chapter 4. Managing Red Hat Cluster With CongaThis chapter describes various administrative tasks for managing a Red Hat Cluster and consists ofthe fo
For Rest art t h is clu st er and St o p t h is clu st er/St art t h is clu st er — Displays apage with the list of nodes for the cluster.For Delet
4. Clicking Go causes a progress page to be displayed. When the action is complete, a page isdisplayed showing the list of nodes for the cluster.4.3.
Rest art t h is service and St o p t h is service — These selections are available when theservice is running. Select either function and click Go t
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Chapter 5. Configuring Red Hat Cluster With system-config-clusterThis chapter describes how to configure Red Hat Cluster software using system-config-
8. Propagating the configuration file to the other nodes in the cluster.Refer to Section 5.9, “ Propagating The Configuration File: New Cluster” .9.
dialog box provides a text box for cluster name and the following checkboxes: Cu st o mCo n f ig u re Mu lt icast and Use a Q u o ru m Disk. In most
ImportantQuorum-disk parameters and heuristics depend on the site environment and specialrequirements needed. To understand the use of quorum-disk par
Fig u re 5.2. Creat in g A N ew Co n f ig u rat io n4. When you have completed entering the cluster name and other parameters in the NewConfiguratio
Fig u re 5.3. T h e Clu st er C o n f ig u rat io n T o o lT ab le 5.1. Q u o ru m- D isk Paramet ersParamet er Descrip t io nUse a Q u o ru m Disk E
Lab el Specifies the quorum disk label created by the mkqdisk utility. If this fieldcontains an entry, the label overrides the Device field. If this f
NoteFor more information about Po st - Jo in Delay and Po st - Fail Delay, refer to thefenced(8) man page.6. Save cluster configuration changes by s
5.5. Adding and Delet ing MembersThe procedure to add a member to a cluster varies depending on whether the cluster is a newly-configured cluster or a
6. Configure fencing for the node:a. Click the node that you added in the previous step.b. At the bottom of the right frame (below Pro p ert ies),
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Section 5.5.1, “Adding a Member to a Cluster”.2. Click Send to Cluster to propagate the updated configuration to other running nodes inthe cluster.3.
d. service rgmanager start5. Start the Red Hat Cluster Suite management GUI. At the Clu st er C o n f ig u rat io n T o o l tab,verify that the con
b. service gfs stop, if you are using Red Hat GFSc. service clvmd stop, if CLVM has been used to create clustered volumesd. service cman stop5. S
5.6. Configuring a Failover DomainA failover domain is a named subset of cluster nodes that are eligible to run a cluster service in theevent of a nod
Section 5.6.3, “ Removing a Member from a Failover Domain”5.6.1. Adding a Failover DomainTo add a failover domain, follow these steps:1. At the left
5. To restrict failover to members in this failover domain, click (check) the Rest rict Failo ver T oT h is Do main s Members checkbox. (With Rest r
configuration.5.6.2. Removing a Failover DomainTo remove a failover domain, follow these steps:1. At the left frame of the Clu st er Co n f ig u rat
the change immediately, choose File => Save to save the changes to the clusterconfiguration.5.7. Adding Clust er ResourcesTo specify a resource for
Fig u re 5.9 . Add in g a Clu st er Service4. If you want to restrict the members on which this cluster service is able to run, choose afailover dom
NoteCircumstances that require enabling Ru n Exclu sive are rare. Enabling Ru nExclu sive can render a service offline if the node it is running on f
IntroductionThis document provides information about installing, configuring and managing Red Hat Clustercomponents. Red Hat Cluster components are pa
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. The f
1. service cman start2. service clvmd start, if CLVM has been used to create clustered volumesNoteShared storage for use in Red Hat Cluster Suite re
Chapter 6. Managing Red Hat Cluster With system-config-clusterThis chapter describes various administrative tasks for managing a Red Hat Cluster and c
Fig u re 6 .1. C lu st er St at u s T o o lYou can use the Clu st er St at u s T o o l to enable, disable, restart, or relocate a high-availabilityser
The following tables describe the members and services status information displayed by the Clu st erSt at u s T o o l.T ab le 6 .1. Memb ers St at u s
ImportantAlthough the Clu st er C o n f ig u rat io n T o o l provides a Q u o ru m Vo t es parameter in theProperties dialog box of each cluster mem
If a cluster member becomes inoperable because of misconfiguration, restore the configuration fileaccording to the following steps:1. At the Clu st e
So, for example, if you have a database that uses an ext3-formatted filesystem, you can disable thedatabase while preserving the filesystem resource f
# chkconfig --level 2345 rgmanager off# chkconfig --level 2345 gfs off# chkconfig --level 2345 clvmd off# chkconfig --level 2345 cman offOnce the prob
Example of Setting Up Apache HTTP ServerThis appendix provides an example of setting up a highly available Apache HTTP Server on a RedHat Cluster. The
Four typographic conventions are used to call attention to specific words and phrases. Theseconventions, and the circumstances they apply to, are as f
2. Use the mkfs command to create an ext3 file system on the partition you created in theprevious step. Specify the drive letter and the partition nu
<Directory /mnt/httpdservice/cgi-bin">AllowOverride NoneOptions None Order allow,deny Allow from all </Directory>Additional changes m
Make sure that the Mo n it o r Lin k checkbox is left checked.Click OK.4. Click the Services property.5. Create the Apache HTTP Server service.Click
Fence Device ParametersThis appendix provides tables with parameter descriptions of fence devices.NoteThe Name parameter for a fence device specifies
SNMP version The SNMP version to use (1, 2c, 3); the default value is 1.SNMP community The SNMP community string; the default value is private.SNMP se
IP address orhostnameThe IP address or hostname assigned to the device.UDP/TCP port(optional)The UD P/TCP port to use for connection with the device;
Module Name (optional) The module name for the DRAC when you have multiple DRACmodules.PasswordScript(optional)The script that supplies a password for
Field Descrip t io nName A name for the server with HP iLO support.Hostname The hostname assigned to the device.Login The login name used to access th
Field Descrip t io nName A name for the IBM iPD U device connected to the cluster into which thefence daemon logs via the SNMP protocol.IP Address The
Password Script(optional)The script that supplies a password for access to the fence device. Usingthis supersedes the Passwo rd parameter.SNMP versio
Kommentare zu diesen Handbüchern