Netbackup Veritas Cluster Implementation Procedure
Install VCS
Receive the VCS license key from Veritas
USER ID: admin
Password: password
Setting environment variables in root /.profile
PATH=$PATH:/sbin:/opt/VRTSvcs/bin:/opt/VRTSllt
Export PATH
MANPATH=$MANPATH:/opt/VRTS/man
Cluster information
Cluster name: bknbha1
Cluster ID: 1
Hostnames: bknbmst1, bknbmst2
Enable the remote copy function between the two servers for VCS installation.
Vi /.rhosts to include the two nodes(bknbmst1, bknbmst2)
Installvcs(only on one sever but you need to setup .rhosts on both servers
Usnb2pd1:/pub/veritas/products/fst_ha.[Link]/cluster_server/installvcs
No SMTP and SNMP
Web-based cluster manager
Verify the following items
/etc/llttab file
/etc/llthosts file (contain every system of the cluster)
/etc/sysname file
/etc/gabtab /sbin/gabconfig –c –n 2
lltstat –nvv
lltstat –l
lltstat –c
/sbin/gabconfig –a
Make sure the /etc/llthosts on both nodes contains the following entries
0 bknbmst1
1 bknbmst2
Verify the /etc/llttab on two nodes
Install Oracle Agent
Ensure the following files are consistent on two cluster nodes
/etc/passwd
/etc/group
/etc/services
/etc/system
/etc/hosts
Pkgadd –d . VRTSvcsor on both
nodes(usnb2pd1:/repo/veritas/dbed_ora_ha.[Link]/oracle_agent/pkgs/VRTSvcsor
Haconf –makerw
Hastop –all –force on the master node
Cp /etc/VRTSvcs/conf/[Link] /etc/VRTSvcs/conf/config
Vi /etc/VRTSvcs/conf/config/[Link] to include the line below
Include “[Link]”
Hastart (master)
Hastart –stale(standby)
Haconf –dump -makero
Create Service group(nbuorasg)
Haconf -makerw
Hagrp –add nbuorasg
Hagrp –modify nbuorasg SystemList bkswmst1 1 bkswmst2 2
Hagrp –modify nbuorasg AutoStartList bkswmst1 bkswmst2
Hagrp –modify nbuorasg bkswmst1 bkswmst2
Hagrp –display nbuorasg
Haconf -dump
Modify it when testing preonline
PreOnline global 1
PreonlineTimeout global 300
Create Resources
Create logical volume group resource
Hares –add dbdg DiskGroup dbdg
Hares –modify dbdg Critical 0
Hares –modify dbdg DiskGroup dbdg
Hares –modify dbdg Enabled 1
Hares –online dbdg –sys bkswmst1
MultiNIC:
1) modify device (local) usnb3dv1, usnb3dv2 two public interfaces
Create logical volume
Hares –add lvudbms LVMLogicalvolume nbuorasg
Hares –modify lvdbms Critical 0
Hares –modify lvdbms LogicalVolume lvdbms
Hares –modify lvdbms VolumeGroup nbuorasg
Hares –modify lvdbms Enabled 1
Hares –modify online –sys bkswmst1
Create mount resources
Hares –add mntdbms Mount dbdg
Hares –modify mntdbms Critical 0
Hares –modify mntdbms BlockDevice \
/dev/vx/dsk/dbdg/lvdbms
Hares –modify mntdbms FSType vxfs
Hares –modify mntdbms MountPoint /DBMS
Hares –modify mntdbms Enabled 1
Hares –online mntdbms –sys bkswmst1
Command
Hacf –verify /etc/VRTSvcs/conf/config/[Link] verify the syntax
Haclus –display view cluster attributes
Hasys –force node1 bring the node from “STALE_ADMIN_WAIT” to running state
Hares –link parent_resource child_resource
Uadmin 1 1 power cycle the nodes
Test
Stop/start VCS on all nodes and leave application running
Hastop –all –force stop VCS on all nodes and leave application running
Hastart –stale start VCS on node 1
Hastart –stale login to node2 and start VCS on node2
Stop/start VCS on local node and leave application running
Hastop –local –force stop VCS on the local node and leave application run
Hastart start VCS on local node
Modify /etc/VRTSvcs/conf/config/[Link]
Hastop –all –force
Vi [Link]
Hastart –force force to read the local [Link]
Rsh node2
Hastart
Haconf –dump -makero
Or
Hastop –all
Vi [Link]
Hastart on primary node
Hastart on standby node
Hagrp –switch nbuorasg –to node2
Ora1 resource
Add the second monitor
user = v3r1tas [Reply:] user
password=gungh0 [Reply:] pword
default tablespace=misc
temporary tablespace=temp
table name =v3r1_date [Reply:] table
monscript:/opt/VRTSvcs/bin/Oracle/[Link]
copy /opt/VRTSvcs/bin/Netbackup/* to new servers
copy /opt/VRTSvcs/bin/triggers/* to both new servers.
Make sure webip has IPmultiNIC too
Add low-priority heartbeat onto public network for BOTH servers
Vi /etc/llttab to add low priority on both machines
link-lowpri qfe:1 /dev/qfe:1
exclude 2-31
hastop –all –force any one
gabconfig –U (unlocad gab)
lltconfig –U (unload llt)
lltconfig –c
gabconfig –c –n2
gabconfig –a (check the seed and cluster status)
hastart on both nodes
gabconfig -a