IBM Support

Changing host and cluster IP on IBM PureData Systems for Analytics Appliance

Troubleshooting


Problem

This is a how-to document that provides step-by-step instructions for changing the cluster IP and host IP's on an IBM PureData Systems for Analytics appliance.

Symptom

Customer wants to change IP address for hosts and cluster IP.

Resolving The Problem

=========================

1. open 2 windows: both logged in as root, directly to ha1 and ha2 (not the wall ip)

2. SSH to ha1 from ha2

3. Shut down the Netezza software by running the following command:

# crm_resource –r nps –p target_role –v stopped

4. Gather current host information

cat /etc/hosts | more

copy down info

5. Verify the port used:

ifconfig -a

take output and match the port with the IP addresses the customer wants to change.

6. Stop the interface by running the following command:

# ifdown bond2

(note: bond2 is just an example, compare cat /etc/hosts and ifconfig -a output to make sure you are on the right port)

7. Open the network configuration file for editing and change the GATEWAY declaration:

# vi /etc/sysconfig/network


8. Open the following file for editing and update it with the new IP address and netmask address:

# vi /etc/sysconfig/network-scripts/ifcfg-bond2

9. Modify /etc/hosts and change the entry for HA1 to reflect the new IP's (for all IP addresses being changed)

vi /etc/hosts

10. Restart the bond2 interface on HA1.

# ifup bond2


11. **** At this point the customer can move the cables and make whatever changes to the switch ports and dns settings for ha1.

Once these changes have been made ...


12. As a root user, log into the HA1 console. ssh to HA2.


13. Stop the bond2 interface by running the following command:

# ifdown bond2

14. Open the network configuration file for editing and change the GATEWAY declaration:

# vi /etc/sysconfig/network

15. Open the following file for editing and update it with the new IP address and netmask address:

# vi /etc/sysconfig/network-scripts/ifconfig-bond2

16. Modify /etc/hosts and change the entry for HA2 to reflect the new IP's (for all IP addresses being changed)

vi /etc/hosts

17. Restart the bond2 interface on HA2.

# ifup bond2


18. **** At this point the customer can move the cables and make whatever changes to the switch ports and dns settings for ha2.

Once these changes have been made make sure that everything is pingable that you changed. If not then the customer may need to engage their network team to find out why as long as what we changed has been done correctly.

Once verified pingable...

19. Change the wall IP address by running the following commands:

a) cd /nzlocal/scripts

b) ./heartbeat_admin.sh –-change-wall-ip –-ip <new wall ip> --mask <new netmask> --dev bond2

c) cibadmin -Q | grep <new wall ip>

21. As a root user, log into the HA1 console.

22. Start the Netezza software by running the following command:

# crm_resource -r nps -p target_role -v started

23. Wait for the cluster (heartbeat) and NPS to come online.

# crm_mon -i2

24. Ping the gateway IP address. If it doesn't resolve, escalate the issue to the customer's Network group.

25. List the peers known to the NTP server:

# ntpq -p

26. Verify hostname resolution with the ping command.

[{"Product":{"code":"SSULQD","label":"IBM PureData System"},"Business Unit":{"code":"BU053","label":"Cloud & Data Platform"},"Component":"Host","Platform":[{"code":"PF025","label":"Platform Independent"}],"Version":"1.0.0","Edition":"","Line of Business":{"code":"LOB10","label":"Data and AI"}}]

Document Information

Modified date:
17 October 2019

UID

swg21694796