Netra™ High Availability Suite 3.0 1/08 Foundation Services Cluster Administration GuideTable Of Contents Previous ChapterNext ChapterBook Index


C H A P T E R  1

Using Cluster Administration Tools and Configuration Files

The Netra High Availability (HA) Suite Foundation Services provide tools and configuration files to administer your cluster environment. Throughout this document, references to a cluster imply the use of a cluster running the Foundation Services. For a description of the cluster environment, refer to the Netra High Availability Suite 3.0 1/08 Foundation Services Overview.

This chapter contains the following sections:

Using the Tools

The Foundation Services tools are located in /opt/SUNWcgha/sbin on Solaris systems, and in /opt/sun/sbin on Linux systems. Include the appropriate directory in your PATH environment variable.

To use some of the Netra HA Suite tools, you must log in as superuser. To know whether you need to log in as superuser, see the man page of the Foundation Services tool.

To add the man pages to your path, refer to the Netra High Availability Suite 3.0 1/08 Foundation Services Manual Installation Guide for the Solaris OS.

Using Netra HA Suite Tools for Cluster Administration

TABLE 1-1 lists the Foundation Services tools you can use for cluster administration. The table describes the purpose of the tools and provides links to their man pages. To add the man pages to your path, refer to the Netra High Availability Suite 3.0 1/08 Foundation Services Manual Installation Guide for the Solaris OS.


TABLE 1-1   Netra HA Suite Cluster Administration Tools 
Tool Description
nhadm

  • Verify the cluster configuration:

    nhadm check

    nhadm check installation

    nhadm check configuration

    nhadm check starting

  • Foundation Services post-installation and OS configuration (Linux only)

    nhadm configure

  • Create the file required for patching shared packages:

    nhadm confshare

  • Display information about node configuration:

    nhadm display

  • Copy files from the master node to the vice-master node. Files listed can be passed as an argument or listed in the data-file file.

    nhadm [d data-file] copy [file]

  • List differences between specified nonreplicated files on the master node and vice-master node:

    nhadm synccheck

  • Store differences between specified nonreplicated files on the master node and vice-master node:

    nhadm syncgen

  • List all possible error messages returned by the nhadm command:

    nhadm -z

  • Display current help options:

    nhadm -h

  • Provide detailed debugging information:

    nhadm -v

  • Display only error messages:

    nhadm -s

For more information, see the nhamd1M man page on the Solaris OS or the nhamd8 man page on Linux.
nhcmmqualif Qualify a node as the master node if the node is master-eligible, and if the cluster has no master node.

For more information, see the nhcmmqualif1M man page on the Solaris OS or the nhcmmqualif8 man page on Linux.
nhcmmrole

  • Get the role of the node. A node can have the following roles:

    MASTER

    VICEMASTER

    IN_CLUSTER

    OUT_OF_CLUSTER

  • List the role of the current node:

    nhcmmrole -v

  • Determine the time out period in seconds > 0:

    nhcmmrole -t

For more information about roles, see the Netra High Availability Suite 3.0 1/08 Foundation Services Glossary and the nhcmmrole1M man page on the Solaris OS or the nhcmmrole8 man page on Linux.
nhcmmstat

  • Analyze the state of a node:

    info

    local

    master

    mynode

    potential

    vice

  • Analyze the state of a group of nodes:

    all

    count

  • Force a reload of the cluster_nodes_table configuration:

    reload

  • Trigger a switchover:

    so

  • Trigger a switchover, from a master node only, switching to a view-master role:

    * so

  • Qualify a node:

    squalif

  • Exit:

    exit

    quit

  • Display help information:

    help

For more information, see the nhcmmstat1M man page on the Solaris OS or the nhcmmstat8 man page on Linux.
nhcrfsadm Authorize or refuse permission to start replication of the master node to the vice-master node.

For more information, see the nhcrfsadm1M man page on the Solaris OS or the nhcrfsadm8 man page on Linux.
nhenablesync Trigger disk synchronization. The nhenablesync tool has no effect when it is used in a shared disk configuration.

For more information, see the nhenablesync1M man page on the Solaris OS or the nhenablesync8 man page on Linux.
nhpmdadm Administer the Daemon Monitor.

  • List all available tags:

    nhpmdadm -L

  • List the individual Netra HA Suite status:

    nhpmdadm -l <name>

For more information about the Daemon Monitor, see the nhpmd1M man page on the Solaris OS or the nhpmd1M man page on Linux.

Using Solaris Tools for Cluster Administration

The following table lists Solaris tools that you can use for cluster administration. The table describes the purpose of the tools and provides links to their man pages. To add the man pages to your path, refer to the Netra High Availability Suite 3.0 1/08 Foundation Services Manual Installation Guide for the Solaris OS.


TABLE 1-2   Solaris Tools Used for Cluster Administration 
Tool Description
boot1M Boot a specific node. For an example of using the boot command, see To Restart a Cluster.
format1M Format, label, repair, and analyze disks on your system. For more information about using the format utility on a cluster, see Using the format Utility to Display and Modify the Configuration of a Disk Partition on a Solaris Node.
ifconfig1M Check and set a network interface configuration, as described in To Obtain Configuration Information About the Network Interfaces of a Node.
init1M Create processes from information stored in the /etc/inittab file.

Do not use this command to reboot or shut down a node. It can be used only to stop and restart Netra HA Suite by reaching the single-user state and then going back to the normal state. To reboot or shut down a node running Netra HA Suite, see General Rules for Shutting Down a Node.
mount1M Mount a file system. For an example of using the mount command, see To Increase the Size of a Replicated Data Partition on a Virtual Disk on a Solaris Node.
netstat1M Show the network status as described in Examining the Routes on a Node. Useful options include netstat -rn for details about routing tables.
patchadd1M Add a patch, as described in Chapter 7.
pgrep1 Find processes by name and other attributes. For an example of using the pgrep command, see To Verify That an nhcmmd Daemon Is Running on Each Peer Node.
pkill1M Kill processes. For an example of using the pkill command, see To Replace Ethernet Cards on a Diskless Node With the DHCP Static Boot Policy.
ps1 List processes. For a list of monitored daemons, see the nhpmd1M man page.
reboot1M Do not use this command to reboot a node running the Foundation Services. Instead, refer to General Rules for Shutting Down a Node for information about rebooting a node that is running the Foundation Services.
route1M Add or remove a route, as described in Examining the Routes on a Node.
truss1 Trace system calls and signals.
ufsdump1M Back up an entire file system or selected files within a file system. For an example of using this command, see To Increase the Size of a Replicated Data Partition on a Virtual Disk on a Solaris Node.
ufsrestore1M Restore files from backup media created with the ufsdump command. For an example of using this command, see To Increase the Size of a Replicated Data Partition on a Virtual Disk on a Solaris Node.

Using Linux Tools for Cluster Administration

The following table lists Linux tools that you can use for cluster administration. The table describes the purpose of the tools and provides links to their man pages.


TABLE 1-3   Linux Tools Used for Cluster Administration  
Tool Description
parted 8 A partition table manipulator for Linux. For more information about using the parted utility on a cluster, see Using the parted Utility to Display and Modify the Configuration of a Disk Partition on a Linux Node.
ifconfig8 Check and set a network interface configuration, as described in To Obtain Configuration Information About the Network Interfaces of a Node.
init8 Create processes from information stored in the /etc/inittab file.

Do not use this command to reboot or shutdown a node. It can be used only to stop and restart the Foundation Services by reaching the single-user state and then going back to the normal state. To reboot or shutdown a node running Netra HA Suite, see General Rules for Shutting Down a Node.
mount8 Mount a file system. For an example of using the mount command, see Increasing the Size of a Replicated Data Partition on a Physical Disk on a Linux Node.
netstat8 Show the network status as described in Examining the Routes on a Node. Useful options include netstat -rn for details about routing tables.
rpm8 RPM Package Manager, used to remove, install, or upgrade a package as described in Chapter 7.
pgrep1 Find processes by name and other attributes. For an example of using the pgrep command, see To Verify That an nhcmmd Daemon Is Running on Each Peer Node.
pkill1 Kill processes. For a list of monitored daemons, see the nhpmd1 man page.
reboot8 Reboot a node running the Foundation Services only with the -f option: do not force halt or reboot using shutdown8. Refer to General Rules for Shutting Down a Node for information about rebooting a node that is running the Foundation Services.
route8 Add or remove a route, as described in Examining the Routes on a Node.
strace1 Trace system calls and signals. For information about using this command, see Stopping and Restarting Daemon Monitoring
dump8 Back up an entire file system or selected files within a file system. For an example of using this command, see To Increase the Size of a Replicated Data Partition on a Virtual Disk on a Linux Node.
restore8 Restore files from backups created with the dump command. For an example of using this command, see To Increase the Size of a Replicated Data Partition on a Virtual Disk on a Linux Node.

Using the Foundation Services Configuration Files in the Cluster Environment

The files created during initial cluster configuration can be used during administration to determine the state of the cluster. By default the configuration files are in the /etc/opt/SUNWcgha/ directory on Solaris systems, and /etc/opt/sun/nhas/ on Linux systems. The following table lists the files created during initial cluster configuration.


Note - NMA is not delivered on Linux. NMA configuration files are not needed for Linux systems.



TABLE 1-4   Netra HA Suite Files Created During Initial Cluster Configuration 
Configuration File Description
cluster_nodes_table Contains a list of the peer nodes in a cluster. There is a copy of the file on each master-eligible node. For information, see the cluster_nodes_table4 man page on the Solaris OS or the cluster_nodes_table5 man page on Linux.
nhfs.conf Contains configuration information for the individual Netra HA Suite. There is a copy of this file on each peer node. For information, see the nhfs.conf4 man page on the Solaris OS or the nhfs.conf5 man page on Linux.
nma.notifs.txt Defines types and targets for the trap notifications of the Node Management Agent (NMA) SNMP. For information, see the nma.notifs.txt4 man page on the Solaris OS or the nma.notifs.txt5 man page on Linux.
nma.params.txt Defines the communication, implementation and accessibility configuration of SNMP in the NMA. For information, see the nma.params.txt4 man page on the Solaris OS or the nma.params.txt5 man page on Linux.
nma.properties Lists the properties of the NMA on a node. There is a copy of this file on each peer node. For information, see the nma.properties4 man page on the Solaris OS or the nma.properties5 man page on Linux.
nma.targets.txt Defines the targets to which the NMA sends SNMP traps. For information, see the nma.targets.txt4 man page on the Solaris OS or the nma.targets.txt5 man page on Linux.
nma.security Configures SNMP security for the NMA. For information, see the nma.security4 man page on the Solaris OS or the nma.security5 man page on Linux.

When you reconfigure a cluster, you might need to modify configuration files. Refer to the man page of a configuration file for information about how to change the file.

Using Solaris Configuration Files in the Cluster Environment

This section describes the Solaris configuration files that can help you determine the current configuration of the cluster. For information about how and when to modify a Solaris configuration file, consult the man page for the file and the Solaris documentation set. The following table lists the Solaris configuration files that you can edit during cluster configuration.


TABLE 1-5   Solaris Files Used With the Foundation Services 
Configuration File Description
/etc/bootparams A database of boot parameters. For information, see bootparams4.
/etc/coreadm.conf Parameters for system core file configuration.
/etc/dfs/dfstab Commands for sharing resources across a cluster. This file does not contain commands for partitions mirrored by Reliable NFS. For information, see dfstab4.
/etc/hostname.interface-name Configuration of each network interface on a node. The file contains the host name of the interface specified by interface-name. If you create a network interface, you must create a /etc/hostname.interface-name file for that interface. For information about network interface configuration, see Examining the Network Configuration Files.
/etc/hosts A local database of host names. For information, see the hosts4 man page. For information about network configuration files, see Examining the Network Configuration Files.
/etc/inet/inetd.conf The list of servers that inetd invokes when it receives an Internet request over a socket. For information, see the inetd.conf4 man page.
/etc/inittab The process dispatching control by the init command. This file contains the rc2 and rc3 entries. For information, see the inittab4 man page. For information about the Daemon Monitor, see the nhpmd1M man page. For the Solaris 10 OS and later, see the smf5 man page
/etc/netmasks A list of the network masks used to implement IP subnetting and their associated IP network numbers. For information, see the netmasks4 man page. For information about network configuration, see Chapter 4.
/etc/nodename The local source for the system name. For information, see the nodename4 man page. For information about network configuration files, see Examining the Network Configuration Files.
/etc/notrouter A file to specify that a node does not act as a router. Use this file to protect against route mismatch. This file must be configured on each peer node to ensure that the node is not routable.
/etc/nsswitch.conf A configuration file for the name service that provides the sources of database information and their lookup order. For information, see the nsswitch.conf4 man page. For information about using the naming services, see Using the Naming Services.
/etc/services A list of the services and aliases available through the Internet. The /etc/services file can be changed to modify the Netra HA Suite ports if, for example, an application is already using the ports configured by default at startup. For information, see the services4 man page. For information about prioritizing the Foundation Services, see Using the Naming Services.
/etc/syslog.conf A file used by the system log daemon, syslogd, to forward a system message to the appropriate log files and users. For information, see the syslog.conf4 man page. For information about configuring system log files, see Chapter 2.
/etc/system A file that customizes the operation of the operating system kernel. You should not have to change the /etc/system file. If you do change the /etc/system file, you must reboot the node for the changes to take effect.
/etc/vfstab A table of file system defaults. For information, see vfstab4.
.profile A file to set up an environment for a user at login time. For information about how to use this file, see the profile4 man page.
.rhosts A list of trusted hosts and users. The .rhosts file specifies remote trusted hosts and users. Modify this file to change remote access authorization between peer nodes. For information, see rhosts4.

Using Linux Configuration Files in the Cluster Environment

This section describes the Linux configuration files that can help you determine the current configuration of the cluster. For information about how and when to modify a Linux configuration file, consult the man page for the file and the Linux documentation set.

The following table lists the Linux configuration files that you can edit during cluster configuration.


TABLE 1-6    Linux Files Used With the Foundation Services  
Configuration File Description
/etc/exports An access control list for NFS file systems being exported. This file does not contain the access control list for partitions mirrored by Reliable NFS. For information, see the exports5 man page.
/etc/fstab A table of file system defaults. For information, see fstab5.
/etc/hosts A local database of host names. For information, see the hosts5 man page. For information about network configuration files, see Examining the Network Configuration Files.
/etc/inetd.conf The list of servers that inetd invokes when it receives an Internet request over a socket. For information, see the inetd.conf5 man page.
/etc/inittab The process dispatching control by the init command. This file contains the rc2 and rc3 entries. For information, see the inittab5 man page.
/etc/netmasks A list of the network masks used to implement IP subnetting and their associated IP network numbers. For information, see the netmasks4 man page. For information about network configuration, see Chapter 4.
/etc/network/interface A configuration of each network interface on a node on MontaVista Linux. For information, see the interface5 man page. For information about network interface configuration see Examining the Network Configuration Files.
/etc/nsswitch.conf A configuration file for the name service that provides the sources of database information and their lookup order. For information, see the nsswitch.conf5 man page. For information about using the naming services, see Using the Naming Services.
/etc/services A list of the services and aliases available through the Internet. The /etc/services file can be changed to modify the Foundation Services ports if, for example, an application is already using the ports configured by default at startup. For information, see the services5 man page. For information about prioritizing the Foundation Services, see Using the Naming Services.
/etc/sysconfig/network-scripts/ifcfg-* A configuration of a network interface on Wind River CGL (for example, /etc/sysconfig/network-scripts/ifcfg-eth0)
/etc/syslog.conf A file used by the system log daemon, syslogd, to forward a system message to the appropriate log files and users. For information, see the syslog.conf5 man page. For information about configuring system log files, see Chapter 2.
/etc/xinetd.conf The list of servers that inetd invokes when it receives an Internet request over a socket. For information, see the xinetd.conf5 man page.
.rhosts A list of trusted hosts and users. The .rhosts file specifies remote trusted hosts and users. Modify this file to change remote access authorization between peer nodes.



Netra™ High Availability Suite 3.0 1/08 Foundation Services Cluster Administration Guide819-5235-13Table Of Contents Previous ChapterNext ChapterBook Index


Copyright © 2008, Sun Microsystems, Inc. All rights reserved.