Steps to run after a PEADM installation

  1. If your backup source PE installation includes compilers and you are restoring your primary server on the same operating system that was in use when the backup was created, migrate your compilers to the new primary server. Run the following command on all compilers: puppet agent -t --server_list <NEW_PRIMARY_SERVER_CERTNAME>.
  2.  If your backup source PE installation includes compilers and you are migrating to a new primary server running a different or upgraded operating system, you must forget the previous compilers and add new compilers on the same OS major version, platform, and architecture as the new primary server. Complete the following steps on the restored primary server:
    1. On the primary server, run the following command for each compiler: puppet node purge <COMPILER_CERTNAME>. If any of the compilers included in your installation were legacy compilers (compilers without the PuppetDB service), you must unpin them from the PE Master node group.

    2. On the new primary server, run Puppet to apply the configuration changes: puppet agent -t

    3. Install and configure new compilers.

  3. If Code Manager was enabled on the primary server you are migrating from, complete the following steps on the new primary server:

    1. If necessary, run the following command to generate a new Puppet token for deploying code: puppet-access login
    2. For each code environment you want to deploy, run: puppet code deploy --wait <ENVIRONMENT_NAME>

    3. Alternatively, if you want to deploy all of your PE code environments, you can run the following command: puppet code deploy --wait --all

  4. Run Puppet to apply changes:
    puppet agent -t --no-use_cached_catalog
  5. If you have not configured the dns_alt_names setting in the pe.conf file, you need to redirect your agents to the new primary server. If your agents have a load balancer address in puppet.conf, this step is unnecessary and you can simply make load balancer changes to redirect agent Puppet traffic. If you are not using load balancers, we recommend updating the server_list configuration setting to the new infrastructure using either the puppet_enteprise::profile::agent::server_list class parameter or with the puppet_conf task.
  6. Once your agents are migrated to the new infrastructure, restart pe-orchestration-services on the primary server. This resets all pxp-agent connections.
    1. puppet resource service pe-orchestration-services ensure=stopped
    2. puppet resource service pe-orchestration-services ensure=running

Related information