HP UX B6941-90001 - Manual (Page 158)

158 Chapter 3
File Tree Layouts on the Managed-Node Platforms
File Tree Layout on Solaris Managed Nodes
NFS Cluster Client on Solaris
In addition to the general rule for determining cluster clients described
in the section “Installation Tips for UNIX Managed Nodes” on page 50
there is also one specific rule for Solaris:
Solaris cluster clients (both with and without disks) are those Solaris
systems that have either a
/usr or /opt file system NFS mounted. Their
cluster server is the system to which /usr or /opt is mounted. If the
/opt file system is not already NFS mounted (only /usr NFS mounted),
then ITO mounts local /opt to /opt of the cluster server:
mount :/opt /opt
The ITO Default Operator on Solaris
The ITO default operator, opc_op, owns /home/opc_op as home
directory. By default, the operator uses the Bourne Shell ( /bin/sh ) and
is locked until the passwd(1M) command is executed. User opc_op
belongs to the group opcgrp.
Currently, user opc_op and group opcgrp are only added locally on the
managed node (useradd or groupadd are used). If the managed node is
a Network-Information-Service (NIS or NIS+) client, the ITO
installation checks if user opc_op is already in the NIS database. If so, no
additional user is installed, otherwise, opc_op is added only locally on the
managed node.
Solaris System Resources Adapted by ITO
ITO makes changes in the following system resource files:
❏ /etc/passwd and /etc/shadow - entry for the default ITO operator
❏ /etc/group - group entry for the default ITO operator
❏ /etc/init.d/opcagt - ITO startup/shutdown script
❏ /etc/rc3.d/S99opcagt - file created
❏ /etc/rc0.d/K09opcagt - file created
❏ /etc/rc1.d/K09opcagt - file created
❏ /etc/vfstab - on cluster client only; entry for mount /opt directory
❏ /etc/init.d/ncs - NCS startup script (if not already present)