Move only the secondary (new FQDN)

In this scenario, the original secondary Data Archive is still online and collecting data. Due to limitations in the connection speed and capacity between the primary and secondary Data Archive nodes, move the secondary server to a new computer. As a result, the name and the FQDN of the secondary Data Archive changes.

You will start buffering in this procedure in order to avoid data loss to the new secondary server. Say that this step is completed at time 't'. After the original secondary Data Archive is stopped in the next step, any data that are still in buffers and that were sent to those buffers prior to time 't' will never be visible in the new secondary node. This includes stale data that may still be in buffers from high availability interface nodes that are not currently primary. To avoid data gaps in the new secondary, one should be sure that data from interface nodes are making it to Data Archive in a reasonably short amount of time.

1. Follow the instructions in the Knowledge Base article: What steps to take when changing the PI Server Host name.

2. Drop the original secondary Data Archive from the collective.

3. Install Data Archive with the same version as the primary Data Archive on the new secondary node.

4. On the primary Data Archive, use Collective manager to add the new secondary Data Archive to the collective.

5. Re-initialize any other existing secondary member servers.

For more information on Collective manager, see Buffering and High Availability.