Sun Microsystems Logo
Products and Services
 
Support and Training
 
 

A  B  C  D  E  F  G  H  I  J  K  L  M  N  O  P  Q  R  S  T  U  V  W  X  Y  Z  
 
File Formatscluster_definition.conf(4)


NAME

 cluster_definition.conf - nhinstall configuration file to define the cluster

SYNOPSIS

 cluster_definition.conf

DESCRIPTION

Configure the cluster_definition.conf file to define the nodes of your cluster for the nhinstall tool. The nhinstall tool does not support the installation of dataless nodes.

The templates for the configuration files are contained in the /opt/SUNWcgha/config.standard directory with .template extensions. Copy the configuration files to a local directory on the installation server as follows:

# mkdir config-file-directory
# export NHOME=/opt/SUNWcgha/config.standard
# cd config-file-directory
# cp $NHOME/cluster_definition.conf.template cluster_definition.conf

Note – All the configuration files must be in the same local directory on the installation server.

The cluster_definition.conf file format is ASCII. Comment lines begin with the comment mark (#). Parameters consist of a keyword followed by an equals (=) sign followed by the parameter value, of the form:

Keyword=Value

Within the Value, you can use a slash at the end of a line to indicate that the Value is continued on the following line. You can also add comments within a Value. For example:

PUBLIC_NETWORK=255.255.255.0 \ #This is the netmask value
192.168.0.0 \ #This is the subnet value

The following Keyword and Value parameters are supported:

CLUSTER_ID

The cluster ID used to assign IP addresses to all the nodes in the cluster. This parameter must be a value between 1 and 254. For example:

CLUSTER_ID=250

It is mandatory to define this parameter.

PASSWORD

The password for the superuser set for all nodes in the cluster. Set this password even if you manually install the Solaris operating system on the master-eligible nodes because this password is required by the nhinstall tool to execute the commands that create the diskless environment.

By default, the password is sunrules.

MEN_INTERFACES

The network interfaces used for Carrier Grade Transfer Protocol (CGTP) on master-eligible nodes. The parameter has the following format:

MEN_INTERFACES=nic0 nic1
  • nic0 is the name of the first network interface.

  • nic1 is the name of the second network interface.

It is mandatory to define this parameter.

NMEN_INTERFACES

The network interfaces used for CGTP on diskless nodes. The parameter has the following format:

NMEN_INTERFACES=nic0 nic1
  • nic0 is the name of the first network interface.

  • nic1 is the name of the second network interface.

It is mandatory to define this parameter if diskless nodes are defined.

LOGICAL_SLICE_SUPPORT

Install the volume management feature of the Solaris operating system. Options are YES and NO. The default is NO.

If LOGICAL_SLICE_SUPPORT is set to YES, the volume management feature of the Solaris operating system is configured for managing replicated partitions and their associated bitmap partitions.

The volume management feature must be used for Netra 20 hardware because the disk scanning mechanism does not guarantee that the slot position of the disk provides a unique and reproducible unit number, for example, when some disks are plugged on FC-AL (Fibre Channel-Arbitrated Loop).

If you are using FC-AL disks on Netra 20 servers, you must do one of the following:

  • Set SOLARIS_INSTALL to YES in the env_installation.conf file.

    Set LOGICAL_SLICE_SUPPORT to YES.

    Define a partition with the mount point set to replica by using the SLICE parameter.

    Complete the configuration and run the nhinstall(1m) tool.

  • Set SOLARIS_INSTALL to NO in the env_installation.conf file.

    Set LOGICAL_SLICE_SUPPORT to NO.

    Install the Solaris operating system yourself on the master-eligible nodes configured with the support of the Solaris Volume Manager (Solaris 9) or Solstice DiskSuite 4.2.1 (Solaris 8).

    Create a disk partition with the attribute replica. This partition is reserved for storing the metadevice database.

    Complete the configuration and run the nhinstall(1m) tool.

SERIALIZE_SYNC

Determine how disk synchronization is performed. Values are YES and NO. The default value is NO.

Synchronization is necessary following a switchover or the vice-master booting, or when you request a full replication. In these circumstances, if SERIALIZE_SYNC is set to NO, Reliable NFS starts the synchonization of all slices at the same time. If SERIALIZE_SYNC is set to YES, slices are synchronized one slice at a time. This reduces the network and disk overhead but increases the time it takes for the vice-master to synchronize with the master. During this time, the vice-master is not eligible to take on the role of master.

USE_WDT

Install and configure the Watchdog Timer packages. Options are YES and NO. The default is NO.

Set the USE_WDT parameter to YES only if you are using Netra servers with hardware watchdogs at the LOM level. When this parameter is set to YES, the Watchdog Timer is installed and configured. In this case, you must install LOM packages. You can install these packages using the addon.conf file. For more information, see the addon.conf(4) man page.

Set the USE_WDT parameter to NO if you are using Netra servers with hardware watchdogs at the OPB level or if you do not want to install the watchdog timer. OPB-level hardware watchdogs are monitored by the server's software.

EXTERNAL_ACCESS

Creates a hostname and an external IP address for the node that has the master role. This IP address is the same for either master-eligible node when that node takes on the master role. When a node takes on the master role, this IP address and hostname are assigned to the master node.

The EXTERNAL_ACCESS parameter has the following format:

EXTERNAL_ACCESS=hostname  IP-address  INTERFACE 
  • hostname is the external host name given to the master node.

  • IP-address is the external IP address given with the master node.

  • NIC is the physical network interface to which the IP address is to be associated.

VENDOR_TYPE

List of platform names of diskless nodes to boot. This parameter can have multiple entries. All entries will be used when the nhinstall tool configures the DHCP.

The VENDOR_TYPE parameter enables you to add a string defining the vendor type sent by a diskless node when issuing a DHCP boot request. This enables you to include boards for diskless nodes that are not part of the tested hardware.

The predefined platform names are:

  • SUNW.UltraSPARC-IIi-cEngine

  • SUNW.UltraSPARC-IIi-Netract

  • SUNW.UltraSPARCengine_CP-60

  • SUNW.UltraAX-i2

  • SUNW.NetraCT-410

  • SUNW.NetraCT-810

To add new vendor type:

VENDOR_TYPE=new_vendor_type

USE_CGTP

Install the CGTP on the cluster. Options are YES and NO. The default is YES.

If USE_CGTP is set to NO, the CGTP packages and patches are not installed. Only the first network interface is considered in the definition of NODE. The CGTP_POSTFIX, if specified, is ignored. In this case, you configure a single network link and your cluster network is not redundant.

PATCH_WITH_SMOSSERVICE

Install diskless patches with smosservice. Options are YES and NO. The default is NO.

If PATCH_WITH_SMOSSERVICE is set to YES, the smosservice patch command installs the patches listed in the addon.conf file and installed on the diskless environment after the Solaris installation. This installation corresponds to phase I or S as described in the addon.conf file. The patchadd command installs all patches applied during phase F. The patchadd command is also used to install any temporary patches, that is patches with names prefixed T.

  • Do not use PATCH_WITH_SMOSSERVICE on the Solaris 9 operating system.

  • You can supersede the patch installation method set with PATCH_WITH_SMOSSERVICE on a per patch basis. For more information, see the addon.conf(4) man page.

IDE_SUPPORT

Configure the Sun StorEdge Network Data Replicator (SNDR) to support IDE disks. Options are YES and NO. The default is NO. For example:

IDE_SUPPORT=YES
RESTRICT_RHOSTS

Restrict the master-eligible nodes from connecting to each other remotely. Options are YES and NO. The default is NO.

If RESTRICT_RHOSTS is set to YES, only the installation server can connect remotely to the master-eligible nodes with rsh and rcp commands. The master-eligible nodes cannot connect to each other remotely. Therefore, you will not be able to run nhadm(1m) commands such as synccheck and syncgen on the master-eligible nodes.

If RESTRICT_RHOSTS is set to NO, the master-eligible nodes can connect to each other remotely with the rsh and rcp commands. The installation server can also connect to the master-eligible nodes remotely.

DISKLESS_BOOT_POLICY

Define the boot policy of the diskless nodes based on a boot policy. Options are DHCP_STATIC, DHCP_DYNAMIC, and DHCP_CLIENT_ID. The default is DHCP_DYNAMIC.

To choose the DHCP dynamic boot policy, specify DHCP_DYNAMIC. In this case, the IP addresses are attributed randomly to diskless nodes from a pool of IP addresses.

To choose the DHCP static boot policy, specify DHCP_STATIC. In this case, the IP addresses are attributed according to the Ethernet address of the diskless nodes. If DHCP_STATIC is selected, the MAC0 and MAC1 attributes for diskless nodes must be set in the NODE parameter. Also, in the NODE parameter, the NIC0 and NIC1 values can be set if the default values specified by NMEN_INTERFACE need to be superseded.

To choose the DHCP client ID boot policy, specify DHCP_CLIENT_ID. In this case, the IP addresses are attributed to a diskless node based on its client ID. This address assignment scheme is relevant only for CompactPCI servers.

REPLICATED_DHCP_FILES

Select the location of the DHCP configuration files. Options are YES and NO. The default is YES.

If REPLICATED_DHCP_FILES is set to YES, the DHCP configuration files are located on a replicated partition and are shared by the master-eligible nodes. If REPLICATED_DHCP_FILES is set to NO, the DHCP configuration files are duplicated on a local partition of each master-eligible node.

CLUSTER_NETWORK

Define the class of IP addresses for your cluster network. The cluster network can have IP addresses of any class. The parameter has the following format:

CLUSTER_NETWORK=netmask  nic0-subnet  nic1-subnet  cgtp-subnet
  • netmask is the mask common to all subnets configured in the /etc/netmasks file, for example, 255.255.0.0.

  • nic0-subnet is the subnet of the first network interface, NIC0. This subnet is configured in the /etc/netmasks file, for example, 172.15.0.0.

  • nic1-subnet is the subnet of the second network interface, NIC1. This subnet is configured in the /etc/netmasks file, for example, 172.16.0.0.

  • cgtp-subnet is the subnet of the virtual network interface, cgtp0. This subnet is configured in the /etc/netmasks file, for example, 172.17.0.0.

By default, class C IP addresses are used as follows:

CLUSTER_NETWORK=255.255.255.0 10.clusterid.1.0 10.clusterid.2.0 10.clusterid.3.0

Where clusterid is the value of the CLUSTER_ID parameter.

DISKLESS_TYPE

Define whether the diskless nodes are to be installed in a cluster or as standalone nodes running only CGTP. Options are CLUSTER and STANDALONE. The default is CLUSTER.

If you specify the CLUSTER option, the diskless nodes are installed with all the Foundation Services. The nodes are configured to be part of the cluster.

If you specify the STANDALONE option, the diskless nodes are installed with the CGTP only. The nodes are configured to run as standalone nodes that do not run all the Foundation Services. The two master-eligible nodes are installed with the Foundation Services so that the standalone diskless node can be booted by the master node.

NODE

Define each node. It is mandatory to define this parameter.

There is an entry for each node and each entry has the following format:

NODE=nodeid {MAC0|client-id|-} {MAC1|-} {name|-} {NIC0|-} {NIC1|-} public-name public-IP public-NIC
  • nodeid

    The ID of the node used to define IP addresses for the node. This option is mandatory.

  • MAC0 | client-id | -

    The Ethernet address of the first network interface of the node. This option is mandatory.

    • For master-eligible nodes, this address is required to boot the master-eligible node from the installation server.

    • For a diskless node, whether this value is required depends on the DHCP boot policy used:

      For the DHCP static boot policy, when the DISKLESS_BOOT_POLICY is set to DHCP_STATIC, specify the Ethernet address of the first network interface of the node.

      For the DHCP dynamic boot policy, when the DISKLESS_BOOT_POLICY is set to DHCP_DYNAMIC, the MAC0 address is ignored and you must specify a hyphen (-).

      For the DHCP client ID boot policy, when the DISKLESS_BOOT_POLICY is set to DHCP_CLIENT_ID, specify the client ID as a string in double quotation marks. You can insert a hexadecimal value, for example "/00999:88:05".

      Note – If you want to include the back slash character, you must include it twice.
  • MAC1

    The Ethernet address of the second network interface of the node. This option is mandatory.

    • For master-eligible nodes, the MAC1 address is ignored and you must specify a hyphen (-).

    • For a diskless node, this value is required depending on the DHCP boot policy used:

      For the DHCP static boot policy, when the DISKLESS_BOOT_POLICY is set to DHCP_STATIC, specify the Ethernet address of the second network interface of the node.

      For the DHCP dynamic boot policy, when the DISKLESS_BOOT_POLICY is set to DHCP_DYNAMIC, the MAC1 address is ignored and you must specify a hyphen (-).

      For the DHCP client ID boot policy, when the DISKLESS_BOOT_POLICY is set to DHCP_CLIENT_ID, the MAC1 address is ignored and you must specify a hyphen (-).

  • name

    Name of the node. By default, the names are assigned as follows:

    • For a master-eligible node, MEN-Cclusterid-Nnodeid.

    • For a diskless node, NMEN-Cclusterid-Nnodeid.

      Do not use underscores ("_") when naming a node. For more information, see hosts(4).

  • NIC0

    Name of the first network interface. This parameter can be set for a diskless node with the DHCP static boot policy only. If you are using the DHCP static boot policy and you want to use the default value, which is the first value defined by the NMEN_INTERFACES parameter, specify a hyphen (-).

    For master-eligible nodes and for diskless nodes with other types of DHCP boot policies, the NIC0 address is ignored. In these cases, you must specify a hyphen (-).

  • NIC1

    Name of the second network interface. This parameter can be set for a diskless node with the DHCP static boot policy. If you are using the DHCP static boot policy and you want to use the default value, which is the second value defined by the NMEN_INTERFACES parameter, specify a hyphen (-).

    For master-eligible nodes and for diskless nodes with other types of DHCP boot policies, the NIC0 address is ignored. In these cases, you must specify a hyphen (-).

  • public-name

    Name of the node on the public network different from the name defined with the name parameter. If PUBLIC_NETWORK is not defined, the public-name is ignored.

  • public-ip

    IP address of the node on the public network. If PUBLIC_NETWORK is not defined, the public-ip is ignored.

  • public-nic

    Network interface for the node supporting the public network. This can be either a physical network interface or an alias. If PUBLIC_NETWORK is not defined, the public-nic is ignored.

Example 1:

NODE=10 08:00:20:f9:c5:54 - node10
NODE=20 08:00:20:f9:a8:12 - node20
NODE=30 -							- node30
NODE=40 -							- node40

Example 2:

NODE=10 08:00:20:f9:c5:54 - node10 - - FSNode1 192.168.12.5 hme1:5
NODE=20 08:00:20:f9:a8:12 - node20 - - FSNode2 192.168.12.6 hme1:101
NODE=30 -							- node30
NODE=40 -							- node40

The first two entries must be the IDs for the master-eligible nodes. The nhinstall tool first installs the product on the first master-eligible node defined and then on the second master-eligible node. The remaining entries define the diskless nodes.

To add diskless nodes to a cluster that is already running, add the definitions for the new nodes by using the NODE parameter and run the nhinstall command with the add option. For information on the nhinstall command to add diskless nodes to the cluster, see the nhinstall(1m) man page.

PUBLIC_NETWORK

Define the public network IP addresses and netmasks for the cluster. The parameter has the following format:

PUBLIC_NETWORK=netmask subnet
  • netmask is the mask for the subnet configured in the /etc/netmasks file. For example: 255.255.255.0

  • subnet is the public subnet configured in the /etc/netmasks file. For example: 192.168.0.0

Note – The PUBLIC_NETWORK parameter also configures the network interface of the installation server. Therefore, the SERVER_IP in the env_installation.conf file is an IP address that is part of the same subnet as defined for the PUBLIC_NETWORK.
DEFAULT_ROUTER_IP

Defines the IP address of the default router for the public network. The DEFAULT_ROUTER_IP is set to an IP address.

DEFAULT_ROUTER_IP=IP address

The default is the public IP address of the installation server.

NIC0_POSTFIX, NIC1_POSTFIX, CGTP_POSTFIX

Defines a suffix string for each host name. These definitions can be set to change the suffix for each host name related to a specific network interface. By default the suffixes are:

  • NIC0_POSTFIX=""

    Therefore, by default, the host name for the NIC0 interface has no suffix.

  • NIC1_POSTFIX="-nic1"

  • CGTP_POSTFIX="-cgtp"

If you specify all the suffixes, as in the following example, the nhinstall tool creates two host names in the /etc/hosts file--one host name with the suffix and one host name without the suffix with respect to the NIC0 interface.

For example, if the suffixes are:

NIC0_POSTFIX="-mynic0"
NIC1_POSTFIX="-mynic1"
CGTP_POSTFIX="-mycgtp"

The resulting /etc/hosts file will include the following type of entry for each node:

10.250.1.10 netraMEN1 netraMEN1-mynic0
10.250.1.20 netraMEN2 netraMEN2-mynic0
10.250.1.30 netraDISKLESS1 netraDISKLESS1-mynic0

Note – Do not use underscores ("_") within a suffix. For more information, see hosts(4).
DIRECT_LINK

You can prevent the occurence of two master nodes in one cluster, by connecting the serial ports of the two master-eligible nodes and defining the DIRECT_LINK parameter. By default, this parameter is not configured.

The DIRECT_LINK parameter has the following format:

DIRECT_LINK=MEN1-serial-device   MEN2-serial-device   speed   [heartbeat-in-seconds]
  • MEN1-serial-device is the serial port on the first master-eligible node to use to connect to the second master-eligible node, for example, /dev/ttya

  • MEN2-serial-device is the serial port on the second master-eligible node to use to connect to the second master-eligible node, for example, /dev/ttya

  • speed is the serial line speed. Valid values for speed are 38400, 57600, 76800, and 115200.

  • heartbeat-in-seconds is the frequency of the heartbeat checking the link between the two master-eligible nodes. The default value for heartbeat is 20 seconds.

BITMAP_IN_MEMORY

Define where the scoreboard bitmaps of the shared partitions are stored.

If you choose the option YES, the scoreboard bitmaps are configured to be stored in memory. In this case, changes are written to the disk only when the node is shut down. This provides better performance on the nodes. However, if both master-eligible nodes fail, the disks must be resynchronized.

Alternatively, if you choose the option NO, the scoreboard bitmaps are configured to write changes to the disk at each update.

For compatibility with previous releases, the default is NO:

BITMAP_IN_MEMORY=NO
NFS_USER_DIR_NOAC

Define the NFS noac option for remote mounted directories. The noac option suppresses data and attributes caching.

If you choose YES, the noac option is configured when mounting remote directories. In this case, all data is retrieved from the master node disk. Data and attribute caching is suppressed.

Alternatively, if you choose NO, the noac option is not configured. In this case, data is cached on the local node.

Use the noac option if the impact on performance is acceptable.

For compatibility with previous releases, the default is YES, for example:

NFS_USER_DIR_NOAC=YES
SLICE_SYNC_TYPE

Define how the replicated partitions are synchronized. Options are:

FS

Only blocks that contain data are replicated. Choose this option for faster synchronization.

RAW

All blocks are replicated. Choose this option for slower synchronization.

The default option for SLICE_SYNC_TYPE is FS.

SLICE_SYNC_TYPE=FS
CHECK_REPLICATED_SLICES

Define whether the replicated partition sanity check is activated. Options are YES and NO. The default is NO.

If CHECK_REPLICATED_SLICES is set to YES, the sanity check is activated. If CHECK_REPLICATED_SLICES is set to NO, the sanity check is not activated.

MASTER_LOSS_DETECTION

Define if the absense of a master node in the cluster must be detected by diskless nodes. Options are YES and NO. The default is YES.

If MASTER_LOSS_DETECTION is set to YES, the absense of a master node in the cluster is detected by the diskless nodes and if such an absense occurs, the diskless nodes are rebooted. If MASTER_LOSS_DETECTION is NO, the absense of a master node in the cluster is not detected by diskless nodes.

SLICE

Defines the disk partitioning. There is an entry for every partition on the disk, and each entry has the following format:

SLICE=name size mount-point bitmap option
  • name

    The name of the disk partition. For example, c0t0d0s0. The name can also be a metadevice name if you have already installed the Solaris operating system configured with the volume management feature. However, slice 2 is reserved.

  • size

    The size of the partition in Mbytes. For slice 7, the size can be set to free unless the partition is replicated.

  • mount-point

    The name of the mount point of the partition. If the partition is replicated, the mount point must be set to unnamed.

    If the partition is reserved for metadevice database storage, the mount point must be set to replica. This is because one partition is required to have the replica attribute when LOGICAL_SLICE_SUPPORT is set to YES. See Example 3 in this man page.

  • bitmap

    The name of the replicated partition. If the partition is not replicated, specify a hyphen (-).

  • option

    The mount option.

    For further details about name, size, mount-point, and option, see the Solaris JumpStart documentation.

SLICE=c0t0d0s0 2048 /               -        logging
SLICE=c0t0d0s1 1024 swap            -        -
SLICE=c0t0d0s3 2048 /export         c0t0d0s5 logging
SLICE=c0t0d0s4 2048 /SUNWcgha/local c0t0d0s6 logging
SLICE=c0t0d0s5 3    unnamed         -        -
SLICE=c0t0d0s6 3    unnamed         -        -
SLICE=c0t0d0s7 free /test1          -        logging

The following disk partitions are mandatory:

  • The root partition, /

  • The /SUNWcgha/local partition

  • If diskless nodes are configured, the /export partition

EXPORTED

The directory to be created and exported on the master-eligible node. There is an entry for every directory to be exported. Each entry has the following format:

EXPORTED=name

name is the directory to be exported. However, do not specify the following directories:

  • /SUNWcgha/local/export because the nhinstall tool automatically creates this directory

  • /export if a diskless environment is required, the nhinstall tool automatically creates this directory

Note – Exported directories must be on a replicated partition because these directories must be accessible regardless of the node that is master. Exporting a non-replicated directory results in errors during a switchover.
MOUNTED

The mount point on the master-eligible node of the directory that is mounted on the master-eligible nodes and the diskless nodes. There is an entry for each mount point, and each entry has the following format:

MOUNTED=mounting-point  remote-dir
  • mounting-point

    The mount point on each node.

  • remote-dir

    The directory name on the master-eligible node.

The following directories are automatically created by the nhinstall tool. Do not define them.

  • The /SUNWcgha/remote directory, which is the mount point for the men_name:/SUNWcgha/local/export/data directory.

  • The /SUNWcgha/services directory, which is the mount point for the men_name:/SUNWcgha/local/export/services/ha_v1/opt.

  • The /SUNWcgha/swdb directory, which is the mount point for the men_name:/SUNWcgha/local/export/services directory.

Where men_name is the host name of the master-eligible nodes.

For remote directories, define only child directories that were previously exported, that is, directories on a replicated partition.

Do not define /export or its subdirectories because it will conflict with the management of diskless nodes.

DATA_MGT_POLICY

Define how the cluster behaves when the vice-master node starts up while the master node is down. Options are:

INTEGRITY

The vice-master waits for the old master to rejoin the cluster before it takes the master role. This ensures that the cluster uses the most up-to-date data.

AVAILABILITY

The vice-master does not wait for the old master to rejoin the cluster before it takes the master role. Data written to the master while the vice-master is down is lost.

ADAPTABILITY

The vice-master checks the disk synchronization state. If the state is not synchronized, that is the state returned by nhcmmstat is synchro:NEEDED, the vice-master waits for the old master to rejoin the cluster. If the state is synchronized, that is the state returned by nhcmmstat is synchro:READY, the vice-master is elected as the new master.

The default value for DATA_MGT_POLICY is INTEGRITY.

DATA_MGT_POLICY=INTEGRITY
SYNC_FLAG

Delays disk synchronization at startup. Options are YES and NO. The default is YES.

If SYNC_FLAG is set to NO, you delay the start of disk synchronization until you use the nhenablesync command. For more information, see the nhenablesync(1m) man page.

MEN_OS_REFERENCE

Defines the profile of the operating system to be used. This parameter is used when the automatic release detection based on the /etc/release file does not enable you to determine the correct operating system version where the Solaris distribution has been updated but there is no change to the Solaris release identification information. Current options are:

S8U7

Solaris 8 2/02 s28s_u7wos_08a SPARC

S8U7_108528-21

Solaris 8 2/02 s28_uwos_08a SPARC -- patched version of the Solaris operating system which is reserved for use with Netra CT 820 hardware

S8PSR3

Solaris 8 HW 7/03 s28s_hw3wos_05a SPARC

S9HWPL3

Solaris 9 s9_58shwp13 SPARC

S9U1

Solaris 9 9/02 s9s_u1wos_08b SPARC

There is no default value for MEN_OS_REFERENCE. The option is automatically detected based on the contents of the /etc/release file.

DISKLESS_OS_REFERENCE

Defines the profile of the Solaris operating system used in the diskless node environment. This parameter is used when automatic release detection based on the contents of the /etc/release file does not permit nhinstall to determine the correct version of the operating system. This can occur when the Solaris distribution has been updated but the release identification information has not changed. Options are:

S8U7

Solaris 8 2/02 s28s_u7wos_08a SPARC

S8U7_108528-21

Solaris 8 2/02 s28_uwos_08a SPARC -- patched version of the Solaris operating system which is reserved for use with Netra CT 820 hardware

S8PSR3

Solaris 8 HW 7/03 s28s_hw3wos_05a SPARC

S9HWPL3

Solaris 9 s9_58shwp13 SPARC

S9U1

Solaris 9 9/02 s9s_u1wos_08b SPARC

There is no default value for MEN_OS_REFERENCE. The option is automatically detected based on the contents of the /etc/release file.

EXAMPLES

Example 1. Example of a standard configuration for a cluster with two master-eligible nodes and one diskless node
#
# Copyright 2003 Sun Microsystems, Inc. All rights reserved.
#
#
# The usage of shell variables is authorized.
# This file is sourced by the nhinstall script.
#

#----------------------------------------------------
#
#  This file enables you to define the cluster
#  environment.
#
#----------------------------------------------------

CLUSTER_ID=250

ARCH=sun4u

MEN_INTERFACES=hme0 hme1

NMEN_INTERFACES=eri0 eri1

LOGICAL_SLICE_SUPPORT=NO

USE_CGTP=YES

RESTRICT_RHOSTS=NO

NODE=10 08:00:20:f9:b3:6a
NODE=20 08:00:20:f9:aa:66
NODE=30 -							- node30

SLICE=c0t0d0s0 2048 /               -        logging
SLICE=c0t0d0s1 1024 swap            -        -
SLICE=c0t0d0s3 2048 /export         c0t0d0s5 logging
SLICE=c0t0d0s4 2048 /SUNWcgha/local c0t0d0s6 logging
SLICE=c0t0d0s5 3    unnamed         -        -
SLICE=c0t0d0s6 3    unnamed         -        -
SLICE=c0t0d0s7 free /test1          -        logging
Example 2. Example of a configuration for a cluster with no diskless nodes

If you have only two master-eligible nodes in your cluster, the disk layout and disk partitioning are different from those of a cluster with one diskless node or more.

The order of the partitions differs from the order in the preceding example as follows:

  • There is no /export partition. Instead, there are two user partitions--one replicated and one not replicated.

  • There is no bitmap partition associated to the /export partition.

  • The user directory is defined as exported on both partitions, and the mount points on the user directory are accessible from both nodes.

#
# Copyright 2003 Sun Microsystems, Inc. All rights reserved.
#
#
# The usage of shell variables is authorized.
# This file is sourced by the nhinstall script.
#

#----------------------------------------------------
#
#  This file enables you to define the cluster
#  environment.
#
#----------------------------------------------------

CLUSTER_ID=250

ARCH=sun4u

MEN_INTERFACES=hme0 hme1

NODE=10 08:00:20:f9:c5:54
NODE=20 08:00:20:f9:a8:12

SLICE=c0t0d0s0 2048 /               -        logging
SLICE=c0t0d0s1 1024 swap            -        -
SLICE=c0t0d0s3 3    unnamed         -        -
SLICE=c0t0d0s4 3    unnamed         -        -
SLICE=c0t0d0s5 2048 /SUNWcgha/local c0t0d0s3 logging
SLICE=c0t0d0s6 2048 /user1          c0t0d0s4 logging
SLICE=c0t0d0s7 free /test1          -        logging

EXPORTED=/user1/export

MOUNTED=/user1_app /user1/export/app
Example 3. Example of a configuration for a cluster with volume-managed, replicated partitions for FC-AL disks
#
# Copyright 2003 Sun Microsystems, Inc. All rights reserved.
#
#
# The usage of shell variables is authorized.
# This file is sourced by the nhinstall script.
#

#----------------------------------------------------
#
#  This file enables you to define the cluster
#  environment.
#
#----------------------------------------------------

CLUSTER_ID=250

ARCH=sun4u

MEN_INTERFACES=hme0 hme1

NMEN_INTERFACES=eri0 eri1

LOGICAL_SLICE_SUPPORT=YES

NODE=10 08:00:20:f9:c5:54
NODE=20 08:00:20:f9:a8:12
NODE=30 -						- node30

SLICE=c0t0d0s0 2048 /               -        logging
SLICE=c0t0d0s1 1024 swap            -        -
SLICE=c0t0d0s3 2048 /export         c0t0d0s5 logging
SLICE=c0t0d0s4 2048 /SUNWcgha/local c0t0d0s6 logging
SLICE=c0t0d0s5 3    unnamed         -        -
SLICE=c0t0d0s6 3    unnamed         -        -
SLICE=c0t0d0s7 free replica         -        logging

ATTRIBUTES

See attributes(5) for descriptions of the following attributes:

ATTRIBUTE TYPEATTRIBUTE VALUE
ArchitectureSPARC
AvailabilitySUNWnhins

SEE ALSO

nhinstall(1M), nhadm(1M), nhenablesync(1M), addon.conf(4), env_installation.conf(4), hosts(4), Solaris Installation Guide


Netra HAS FS 2.1Go To TopLast Changed September 2004