The unit that VERITAS Cluster Server uses for switching systems between nodes is called a group. The applications comprising a group operate using what are referred to as resources. The following types of resources are available:

• Disk group (shared disk)
• Network address (logical IP address)
• Network interface card (NIC)
• Service (application)

(a) Setting up groups and resources

Provide NICs, logical IP addresses, and a shared disk, and set them up as VERITAS Cluster Server resources to form a group. In the explanation in this manual, VERITAS Volume Manager is used to set up a shared disk. When these resources are set up, resource types that have already been defined (such as NIC type, IP type, and DiskGroup) can be used.

You also need to set up HiRDB as a resource within the group so that it can be managed by VERITAS Cluster Server. For this purpose, define a new resource type for HiRDB. For details about how to define a resource type, see 26.2.7(2) HiRDB resource type definition.

The resource type name is HiRDB_S for a HiRDB single server configuration and HiRDB_P for a HiRDB parallel server configuration. This manual uses the generic notation HiRDB_x for resource type names. Substitute HiRDB_S or HiRDB_P as appropriate.

(b) Defining the parent-child relationships of resources

Parent-child relationships are defined for the resources defined within a group. For HiRDB to run, logical IP addresses (when IP addresses are inherited) and the shared disk must be enabled. Therefore, HiRDB_x type resources must be defined as parent resources to the IP type and DiskGroup type resources. The following figure shows a group configuration.

Figure 26-41 Group configuration

(c) Dummy file (applicable to the monitor mode only)

If HiRDB terminates abnormally when the system switchover facility is operating in the monitor mode, HiRDB will restart. For this reason, it is not necessary to monitor the operating status of HiRDB. However, you must create a dummy file for notifying VERITAS Cluster Server that HiRDB is currently active as a resource. This dummy file must satisfy all of the following conditions.

Conditions

• The dummy file is created when HiRDB is started by VERITAS Cluster Server.
• The dummy file is deleted when HiRDB is terminated by VERITAS Cluster Server.
• While the dummy file is in existence, the HiRDB resource is considered to be running.
  

  Hint

  If the dummy file is deleted inadvertently, VERITAS Cluster Server assumes that an error has occurred in the resource and switches systems. To prevent this, specify Critical=0 as the resource attribute value for the HiRDB_x type resource. Also, to prevent the dummy file from being deleted inadvertently, create it using the name $PDDIR/.pdveritas.

(d) Notes

• The standard output messages of HiRDB commands executed from the agent script of a HiRDB_x type resource are output to VERITAS Cluster Server's log file (/var/VRTSvcs/log/engine.log_A).
• Do not delete the dummy file for the HiRDB_x type resource. If it is deleted, system monitoring cannot be performed correctly.

To set up HiRDB as a resource, you must define resource type HiRDB_x for HiRDB. When you create a new resource type, you must also define an agent to monitor the resource. For details about how to define agents, see 26.2.7(3) Agent definition pre-preparation.

(a) HiRDB single server configuration

The following example shows a resource type definition for a HiRDB single server configuration:

 
type HiRDB_S (
     static str ArgList[] = { PdDir, PdConfPath, Ld_Library_Path, DummyFilePath }
     str PdDir
     str PdConfPath
     str Ld_Library_Path
     str DummyFilePath
)
 

Create a file containing the above definition under the name /etc/VRTSvcs/conf/config/HiRDB_STypes.cf.

(b) HiRDB parallel server configuration

The following example shows a resource type definition for a HiRDB parallel server configuration:

 
type HiRDB_P (
     static str ArgList[] = { PdDir, PdConfPath, Ld_Library_Path, DummyFilePath }
     str PdDir
     str PdConfPath
     str Ld_Library_Path
     str DummyFilePath
)
 

Create a file containing the above definition under the name /etc/VRTSvcs/conf/config/HiRDB_PTypes.cf.

Define an agent for the newly-created resource type HiRDB_x. This subsection explains how to define an agent using a shell script. Before defining an agent, take the following pre-preparation step.

• HiRDB single server configuration
  Copy /opt/VRTSvcs/bin/ScriptAgent under the name /opt/VRTSvcs/bin/HiRDB_S/HiRDB-SAgent.
  
• HiRDB parallel server configuration
  Copy /opt/VRTSvcs/bin/ScriptAgent under the name /opt/VRTSvcs/bin/HiRDB_P/HiRDB-PAgent.
  

Define the action details of the agents as listed in the following table.

Table 26-14 Agent action definition items and file names

(a) Online script

An online script explains the details of the processing to be performed when an agent brings a resource online. The following processing is required:

• Setting of the environment variables necessary for executing $PDDIR/bin/pdstart
• HiRDB startup
• Dummy file creation
• Dummy file mode change

HiRDB single server configuration

The following example shows an online script for a HiRDB single server configuration:

 
#!/bin/sh
PATH=/sbin:/usr/bin:/usr/sbin:/etc:/bin:/opt/VRTSvcs/bin:"$2"/bin
export PATH
PDDIR="$2"
PDCONFPATH="$3"
LD_LIBRARY_PATH="$4"
export PDDIR PDCONFPATH LD_LIBRARY_PATH
$PDDIR/bin/pdstart
/bin/touch "$5"
/bin/chmod 0400 "$5"
 

HiRDB parallel server configuration

The following example shows an online script for a HiRDB parallel server configuration:

 
#!/bin/sh
PATH=/sbin:/usr/bin:/usr/sbin:/etc:/bin:/opt/VRTSvcs/bin:"$2"/bin
export PATH
PDDIR="$2"
PDCONFPATH="$3"
LD_LIBRARY_PATH="$4"
export PDDIR PDCONFPATH LD_LIBRARY_PATH
$PDDIR/bin/pdstart -q
/bin/touch "$5"
/bin/chmod 0400 "$5"
 

Note

The pdstart -q command starts the units in a HiRDB parallel server configuration during use of the system switchover facility.

(b) Offline script

An offline script explains the details of the processing to be performed when an agent takes a resource offline. The following processing is required:

• Setting of the environment variables necessary for executing $PDDIR/bin/pdstop
• HiRDB termination
• Dummy file deletion

HiRDB single server configuration

The following example shows an offline script for a HiRDB single server configuration:

 
#!/bin/sh
PATH=/sbin:/usr/bin:/usr/sbin:/etc:/bin:/opt/VRTSvcs/bin:"$2"/bin
export PATH
PDDIR="$2"
PDCONFPATH="$3"
LD_LIBRARY_PATH="$4"
export PDDIR PDCONFPATH LD_LIBRARY_PATH
$PDDIR/bin/pdstop -f -q
/bin/rm -f "$5"
 

Note

Specify the pdstop -f -q command to forcibly terminate HiRDB.

This offline script is executed when a system switchover occurs. When this script executes, HiRDB is forcibly terminated so that the systems can be switched immediately and the standby system can restart HiRDB in order to resume operations.

HiRDB parallel server configuration

The following example shows an offline script for a HiRDB parallel server configuration.

 
#!/bin/sh
PATH=/sbin:/usr/bin:/usr/sbin:/etc:/bin:/opt/VRTSvcs/bin:"$2"/bin
export PATH
PDDIR="$2"
PDCONFPATH="$3"
LD_LIBRARY_PATH="$4"
export PDDIR PDCONFPATH LD_LIBRARY_PATH
$PDDIR/bin/pdstop -z -q
/bin/rm -f "$5"
 

Note

Specify the pdstop -z -q command to forcibly terminate HiRDB.

This offline script is executed when a system switchover occurs. When this script executes, HiRDB is forcibly terminated so that the systems can be switched immediately and the standby system can restart HiRDB in order to resume operations.

(c) Monitor script

A monitor script explains the details of the processing to be performed when an agent monitors a resource (to check whether a resource is online). The following processing is required:

• Check of whether a dummy file exists
• Return value setting

The following example shows a monitor script:

 
#!/bin/sh
if /bin/test -f "$5"
then
    exit 110
else
    exit 100
fi
 

The values of the environment variables and dummy file path names that are needed in each script are passed as arguments during script execution. For the arguments and return values that are passed to each script, see the VERITAS Cluster Server documentation.

Create the environment setup file for VERITAS Cluster Server (/etc/VRTSvcs/conf/config/main.cf) and set up groups and resources.

(a) Resource attribute setting value

The table below lists the values to be specified for resource attributes. For details about the individual items, see the VERITAS Cluster Server documentation.

Table 26-15 Values to be specified for resource attributes

(b) Environment setup file example (when IP addresses are inherited)

include "types.cf"
include "HiRDB_STypes.cf"
 
cluster vcs (
      UserNames = { vcsadm = cDi1yyJLOgPWY }
      CounterInterval = 5
      Factor = { runque = 5, memory = 1, disk = 10, cpu = 25, network = 5 }
      MaxFactor = { runque = 100, memory = 10, disk = 100, cpu = 100, network = 100 }
      )
 
system mainhost
 
system reservedhost
 
snmp vcs (
        TrapList = { 1 = "A new system has joined the VCS Cluster",
                 2 = "An existing system has changed its state",
                 3 = "A service group has changed its state",
                 4 = "One or more heartbeat links has gone down",
                 5 = "An HA service has done a manual restart",
                 6 = "An HA service has been manually idled",
                 7 = "An HA service has been successfully started" }
        )
 
group gr1 (
        SystemList = { mainhost, reservedhost }
        AutoStartList = { mainhost }
        )
 
        HiRDB_S gr1_HiRDB_S_UNT1 (
                Critical = 0
                PdDir = "/hirdb/pddir_s"
                PdConfPath = "/hirdb/pddir_s/conf"
                Ld_Library_Path = "/hirdb/pddir_s/lib"
                DummyFilePath = "/hirdb/pddir_s/.pdveritas"
                )
 
        DiskGroup gr1_DiskGroup_sharedg1 (
                DiskGroup = sharedg1
                )
 
        IP gr1_IP_logicalhost (
                Device = hme0
                Address = "172.16.161.177"
                )
 
        NIC gr1_NIC_hme0 (
                Device = hme0
                NetworkHosts = { "172.16.161.1" }
                )
 
        gr1_HiRDB_S_UNT1 requires gr1_DiskGroup_sharedg1
        gr1_DiskGroup_sharedg1 requires gr1_IP_logicalhost
        gr1_IP_logicalhost requires gr1_NIC_hme0

(c) Environment setup file example (when IP addresses are not inherited)

include "types.cf"
include "HiRDB_STypes.cf"
 
cluster vcs (
      UserNames = { vcsadm = cDi1yyJLOgPWY }
      CounterInterval = 5
      Factor = { runque = 5, memory = 1, disk = 10, cpu = 25, network = 5 }
      MaxFactor = { runque = 100, memory = 10, disk = 100, cpu = 100, network = 100 }
      )
 
system mainhost
 
system reservedhost
 
snmp vcs (
        TrapList = { 1 = "A new system has joined the VCS Cluster",
                 2 = "An existing system has changed its state",
                 3 = "A service group has changed its state",
                 4 = "One or more heartbeat links has gone down",
                 5 = "An HA service has done a manual restart",
                 6 = "An HA service has been manually idled",
                 7 = "An HA service has been successfully started" }
        )
 
group gr1 (
        SystemList = { mainhost, reservedhost }
        AutoStartList = { mainhost }
        )
 
        HiRDB_S gr1_HiRDB_S_UNT1 (
                Critical = 0
                PdDir = "/hirdb/pddir_s"
                PdConfPath = "/hirdb/pddir_s/conf"
                Ld_Library_Path = "/hirdb/pddir_s/lib"
                DummyFilePath = "/hirdb/pddir_s/.pdveritas"
                )
 
        DiskGroup gr1_DiskGroup_sharedg1 (
                DiskGroup = sharedg1
                )
 
        NIC gr1_NIC_hme0 (
                Device = hme0
                NetworkHosts = { "172.16.161.1" }
                )
 
        gr1_HiRDB_S_UNT1 requires gr1_DiskGroup_sharedg1
        gr1_DiskGroup_sharedg1 requires gr1_NIC_hme0