I want to move my CAPAM hardware (or VMs) to a new datacenter
Article ID: 10010
Updated On:
Products
Issue/Introduction
I am planning to move my CAPAM cluster to a new data center.
Cause
The process could involve other changes that need to be accounted for depending on how complex your cluster and network is setup.
Environment
Resolution
1- Decide whether you are going to be in production or not when you decide to start this process.
Assuming that you will still be in production, leave Primary Node 1 in production, and move Node 2 over to the new data center.
If there are more than two nodes, it would have to be decided if you want to have redundancy in the old datacenter or the new datacenter while CAPAM is being moved.
2- You are still in production and clustered.
Put Node 2 into Maintenance Mode by going to Config/Diagnostic/Maintenance Mode and click the “On” button to Turn On Maintenance Mode.
Then go to Configuration/Clustering and click the Turn Cluster Off button at the bottom of the page.This should stop clustering on all nodes; verify by logging into all nodes separately. (you can stop clustering from either Node 1 one or Node 2 (or any node in that cluster)).
Unlock the database on Node 1 and Node 2, etc.
Users will need to log directly into Node 1 using its IP, the VIP does not work when unclustered. If a cluster is being used with Node1,
3- Save and download the database and config from Node1 as a precaution (config will be useless if are changing your IPs / subnet).
Perform VM Snapshots of the unclustered Nodes as well.
Network reconfiguration can be done for hardware, using the front LCD panel. On VMs, if can be done at the VMWare Console panel using the menus in this screen.
4- Edit Node 2 IP/subnet info on the Configuration/Network page, and edit the IP info on the Configuration/Clustering page (this may not be able to be configured until all nodes are moved).
While on Clustering page, move the Node1 IP down, and move the Node2 IP up to the top. If you are changing all the IPs, then move Node2's new IP the top IP.
5- If you are changing IPs and subnets, you may need to edit your target devices as well.
You can do this manually on the Devices page on the Dashboard. If you have a large amount of target devices, you can facilitate this process by going to Devices/ Import/Export Devices/and click on Export Devices. You can edit this list and import the same list with all your target device IP changes.
6- Go to Configuration/Power and select Power Off or Stop Instance to shutdown Node2.
7- Node 2 has now been moved to the new data center. Connect power cord(s) or power on VM(s) take Node 2 out of Maintenance Mode.
Make sure the unit is accessible on the network, and that you can reach the target devices.
8- Give users the new IP for Node 2 (new Primary), then put Node 1 into Maintenance mode.
9- Change the Configuration/Network info and the Config/Clustering info in the same way that you updated Node2 (with Node2’s IP address at the top in the primary position.
Use Configuration/Power to powerdown Node 1.
10- Node1 is now in the new datacenter.
Connect power cord(s) and power on Node1 (new Secondary), and make sure that it is accessible on the network.
11- On Node 2 (new Primary) go to the Configuration/Clustering page, and starting at the left, click Save Config Locally, then click Save to Cluster, and then click on Turn Cluster On.
12- Once the cluster has formed, which means that the database has been passed between Node2 and Node1, all information is synchronized and Node2 is the new Primary and Node1 is the new Secondary and all targets should be reachable (as long as all the IP and subnet information is correct or has been edited for the new data center).
Additional Information
This document covers everything you need to know for a data center change, but there could be some things I have not covered for your specific configuration.
There could be LDAP servers, NTP servers or any other external IP source such as a recording server, Syslog server or any other server that CAPAM communicates with besides the other node(s) and the target servers.
Feedback
