Modify a network used by high availability

Modify your network settings to move an existing high availability environment to another network. For example, if the environment was built in one location, and then had to be set up at a customer site.

1. Power on the primary management node and change the network settings. The Platform Cluster Manager Community Edition services should be down.
   For Linux, change the IP addresses in /etc/sysconfig/network-scripts/ifcfg-* scripts.

   #service network restart
   
   Use ifconfig to check if the new IP addresses are available, and configure 
   the virtual provision IP address
   #ifconfig eth1:0 192.168.2.10 netmask 255.255.255.0

2. Change the shared file systems on management node to use the new NFS server.

   #mount –t nfs –o vers=3, rw,noacl 192.168.2.100:/export/data/shared_pcm
   #mount –t nfs –o vers=3,rw, noacl 192.168.2.100:/export/home /shared_pcm_home
   #source /etc/profile.d/pcmenv.sh

3. Start database server and xCAT daemon.

   #vi /var/lib/pgsql/data/pg_hba.conf 
   host all all 192.168.1.10/32 md5 
   becomes 
   host all all 192.168.2.10/32 md5
   
   
   #service postgresql start
   #service xcatd start
   #lsxcatd -a

4. Change the site table attributes.

   #chdef -t site pcm_mntp_work=192.168.1.100:/export/data
   #chdef -t site pcm_mntp_home=192.168.1.100:/export/home
   # chdef -t site master=192.168.2.10 
   # chdef -t site nameservers=192.168.2.10

5. Change the NIC table attributes for all management nodes, including the primary management node, the secondary management node, and the virtual management node.

   # chdef -t node -o primarymn nicips='eth0!172.20.8.3,
   eth1!192.168.2.3’ ip=192.168.2.3
   # chdef -t node -o backupmn nicips='eth0!172.20.8.4,
   eth1!192.168.2.4’ ip=192.168.2.4
   # chdef -t node -o virtualmn nicips='eth0:0!172.20.8.10,
   eth1:0!192.168.2.10’ ip=192.168.2.10

6. Update networks table as needed using the tabedit and chtab commands.

   #tabedit
   #tabdump networks
   #netname,net,mask,mgtifname,gateway,dhcpserver,tftpserver,nameservers,
   ntpservers,logservers,dynamicrange,staticrange,staticrangeincrement,
   nodehostname,ddnsdomain,vlanid,domain,comments,disable
   "provision","192.168.2.0","255.255.255.0","eth1","<xcatmaster>",,,,,,
   "192.168.2.201-192.168.2.254","192.168.2.3-192.168.2.200","1",,,,
   "private.dns.zone",,
   "public","172.20.8.0","255.255.255.0","eth0","172.20.8.2",,,,,,,"172.20.8.3
   -172.20.8.10","1",,,,,,

7. Optional: If needed, change the management node name.

   #chdef -t site pcm_master_node=’newprimarymn,newbackupmn|newvirtualmn’
   #chdef -t node -o primarymn -n newprimarymn
   #chdef -t node -o backupmn -n newbackupmn
   #chdef -t node -o virtualmn -n newvirtualmn
   
   #hostname newprimarymn 

8. Reconfigure the PCMHA service agent and other services on primary management node to use the new virtual IP address.

   #pcm-ha-support gendata
   #pcmconfig -f

9. Power on the secondary management node to change its network settings on its console.

   Change IP addresses in /etc/sysconfig/network-scripts/ifcfg-* scripts
   #service network restart
   Use ifconfig to check if the new IP addresses are available
   
   #hostname newbackupmn
   Edit /etc/sysconfig/network and change the HOSTNAME attribute

10. Run below commands on active management node to start Platform Cluster Manager Community Edition services on the standby management node.

    #updatenode newbackupmn -F
    #ssh newbackupmn ‘mount -a’
    #ssh newbackupmn ‘service pcm start’

11. Run the diagnostic tool to check high availability environment.

    #pcmhatool check

12. Regenerate the new IP addresses for existing compute nodes and re-provision them.

    #pcmhatool failmode -m manual
    #noderegenips __Managed
    #nodeset __Managed osimage=’rhels6.4-x86_64-stateful-compute’