Chapter 4
Examples of Configuration for Servicing NFS Jobs

The HA configuration in each network is different. How it should be configured depends on two things:


In this chapter, we provide two examples to illustrate how you should configure HA to service respectively one NFS job
and two NFS jobs.

4.1 Exemplary Configuration for Servicing One NFS Job

  1. Drawing of network connection for one NFS job:


  2. Basic information about the exemplary network machines:

127.0.0.1......localhost
#
# local interfaces
192.9.2.2......................... russell
..............loghost
192.9.3.2..........................russell-1
#
# Other HA server interfaces
192.9.2.1............. peter
192.9.3.1....................... peter-1
#
# HA Active interfaces
192.9.3.9............. act1
#
# PCNFS client
192.9.3.10........... client

The next table provides the information about them that you can see from the "peter" machine in its /etc/hosts.


127.0.0.1....................localhost
#
# local interfaces
192.9.2.1................... peter...... loghost
192.9.3.1....................peter-1
#
# Other HA server interfaces
192.9.2.2................... russell
192.9.3.2................... russell-1
#
# HA Active interfaces
192.9.3.9................... act1
#
# PCNFS client
192.9.3.10................. client

The next table provides the information about them that you can see from the client machine in its /etc/hosts.


127.0.0.1...................localhost
#
192.9.3.10.................... client loghost
# HA Active interfaces
192.9.3.9.......................act1

  1. Exemplary configuration for one nfs job:


# HA configuration file for ONE NFS
#
#################################
#..........License Definition........................ #
#################################
#
LICENSE=russell:6BD84BC01095F57A
LICENSE=peter:EAFEF623F7F970AF
SERIAL=russell:1234567890123456
SERIAL=peter:1234567890123456
DATE=russell:03201996
DATE=peter:03201996
#
#
#################################
#..........Node Definition........................... #
#################################
#
HOST_NODE=russell,peter
#
ORGINAL_IP=russell:le0:russell,le1:russell-1
ORGINAL_IP=peter:le0:peter,le1:peter-1
ORGINAL_ETHER=russell:8:0:20:1a:ab:8e
ORGINAL_ETHER=peter:8:0:20:1b:cd:99
#
#
#################################
#.....................Job Definition................... #
#################################
#
JOB_NAME=nfs1
#
JOB=nfs1
#
ACT_NODE=russell
ACT_LAN=le1
ACT_IP=act1
#
SHARE_DISK=/dev/dsk/c1t3d0s0
MOUNT_POINTER=/share_nfs1
#
STANDBY_NODE=peter
STANDBY_LAN=le1
STANDBY_DISK=/dev/dsk/c1t3d0s0
#
SUPPORT=NFS
#
#
#################################
#..............Misc. Definition....................... #
#################################
#
HEART_BEAT=russell:le0
HEART_BEAT=peter:le0
#
ALIVE_CHECK_TIME=8
DEVICE_CHECK_TIME=10
SYNC_DISK_TIME=5
#
##end

Notes:
  1. To access the job provided on the server, the client machine must first mount
    the partition information on its hard disk.

  2. The line "SUPPORT=NFS" should be in the configuration.
  3. When servicing an NFS job, the major device number and minor device number
    viewed from both the "peter" and "russell" machinese must be the same. To ensure
    the numbers are the same for both machines, place in the same slot position the
    second SCSI cards used to connect both machines to the shared disk.
  4. It is recommended to set in the configuration the SYNC_DISK_TIME at
    5 seconds for an NFS job.
  5. If your standby ethernet interface is not le1 (e.g., qe3 as shown in the following
    figure), you need to change the STANDBY_LAN in the configuration from


Table of ContentsSec. 4-1 sec. 4-2