Upgrading Puppet Enterprise

Upgrade your PE installation as new versions become available.

Upgrade paths

These are the valid upgrade paths for PE.


If you're on version...	Upgrade to...	Notes
2019.8 (latest)	You're up to date!
2019.y	latest
2018.1.2 or later 2018.1.3 or later with disaster recovery	latest	You must have version 2018.1.2 or later in order to complete prerequisites for upgrade to the latest version. With disaster recovery enabled, you must have version 2018.1.3 in order to upgrade to the latest version. Alternatively, you can forget and then recreate your replica after upgrade.
2018.1.0 or 2018.1.1	2018.1.z
2017.3.z	2018.1
2017.2.z	2018.1
2017.1.z	2018.1
2016.5.z	2018.1
2016.4.10 or later	2018.1
2016.4.9 or earlier	2016.4.z, then 2018.1	To upgrade to 2018.1 from 2015.2.z through 2016.4.9, you must first upgrade to the latest 2016.4.z.
2016.2.z
2016.1.z
2015.3.z
2015.2.z
3.8.x	2016.4.z, then 2018.1	To upgrade from 3.8.x, you must first migrate to the latest 2016.4.z. This upgrade requires a different process than upgrades from other versions.

Upgrade cautions

These are the major changes to PE since the last long-term support release, 2018.1. Review these recommendations and plan accordingly before upgrading to this version.

Important: Always back up your installation before performing any upgrade.

Retired primary server platforms in 2019.8

Support for Enterprise Linux 6 and Ubuntu 16.04 as a primary server platform was removed in 2019.8. If your primary server is installed on one of these platforms, you must update the operating system before you can upgrade to this version of PE.

Follow these steps to upgrade from an unsupported primary server platform.

Back up your existing installation.
Configure a new node with a supported primary server platform, for example Enterprise Linux 7 or 8, or Ubuntu 18.04.
Install your current PE version on the new node.
Restore your installation on the new primary server using the backup you created.
Upgrade to the latest PE version.
If you have compilers, reprovision them.
If you have a replica, forget and then reprovision it.

PuppetDB migrations in PE 2019.1, 2019.3, and 2019.7

Deleting PuppetDB reports and truncating the resource events table before you upgrade can reduce migration time and lessens downtime, especially when your upgrade involves a database migration.

The method you use to delete reports depends on the version you're upgrading from:

If you're upgrading from 2019.7 or later, or PE LTS version 2018.1.15 or later, run the packaged delete reports command on your primary server before upgrade: /opt/puppetlabs/bin/puppetdb delete-reports
If you're upgrading from other versions, which don't include the packaged delete reports command, manually delete reports by following these steps.

On your primary server, stop the PuppetDB service: puppet resource service pe-puppetdb ensure=stopped
On your PE-PostgreSQL server, create a file named /tmp/delete-reports.sql and set it to be owned by the pe-postgres user (chown pe-postgres:pe-postgres /tmp/delete-reports.sql).

Add contents to the .sql file according to your PE version.

PE versions earlier than 2019.3

BEGIN TRANSACTION;
                            
                            ALTER TABLE certnames DROP CONSTRAINT IF EXISTS certnames_reports_id_fkey;
                            UPDATE certnames SET latest_report_id = NULL;
                            TRUNCATE TABLE reports CASCADE;
                            
                            ALTER TABLE certnames
                            ADD CONSTRAINT certnames_reports_id_fkey
                            FOREIGN KEY (latest_report_id) REFERENCES reports(id) ON DELETE SET NULL;
                            
                            COMMIT TRANSACTION;

PE 2019.3 through 2019.6

BEGIN TRANSACTION;
                        
                        ALTER TABLE certnames DROP CONSTRAINT IF EXISTS certnames_reports_id_fkey;
                        UPDATE certnames SET latest_report_id = NULL;
                        
                        DO $$ DECLARE
                        r RECORD;
                        BEGIN
                        FOR r IN (SELECT tablename FROM pg_tables WHERE tablename LIKE 'resource_events_%') LOOP
                        EXECUTE 'DROP TABLE ' || quote_ident(r.tablename);
                        END LOOP;
                        END $$;
                        
                        TRUNCATE TABLE reports CASCADE;
                        
                        ALTER TABLE certnames
                        ADD CONSTRAINT certnames_reports_id_fkey
                        FOREIGN KEY (latest_report_id) REFERENCES reports(id) ON DELETE SET NULL;
                        
                        COMMIT TRANSACTION;

Run the command: su - pe-postgres -s /bin/bash -c "/opt/puppetlabs/server/bin/psql -d pe-puppetdb -f /tmp/delete-reports.sql"
Restart the PuppetDB service: puppet resource service pe-puppetdb ensure=running

Java 11 upgrade in PE 2019.3

PE 2019.3 includes an upgrade from Java version 8 to version 11. If you've customized PE Java services, or use plug-ins that include Java code, test PE 2019.3 and later thoroughly in a non-production environment before upgrading.

Retired split installations in PE 2019.2

Split installations, where the primary server, console, and PuppetDB are installed on separate nodes, are no longer supported.

Before upgrading to 2019.2 or later, you must migrate from a split to a standard installation. For instructions, see Migrate from a split to a standard installation in the documentation for your current version.

Orchestrator memory use increase in PE 2019.2

Puppet orchestrator uses more memory in version 2019.2 than in previous versions due to the addition of a Java virtual machine (JVM), which enables new features and functionalities such as plans. If your memory use is near capacity when running PE 2019.1 or older versions, allocate additional memory before upgrading to PE 2019.2.

Additionally, take care when writing plans, as they can require more memory than is allocated to the orchestrator. To work around this issue, rewrite the plan or increase the memory allocated to the orchestrator.

PostgreSQL 11 upgrade in PE 2019.2

PE 2019.2.0 includes an upgrade from pe-postgresql version 9.6 to version 11. This upgrade involves a datastore migration that requires extra disk space (110% of the current 9.6 datastore) and extra time to upgrade (roughly two and four minutes of additional time per 10GB of datastore size). The installer issues a warning and cancels the upgrade if there is insufficient space.

To review the size of your PostgreSQL installation as well as the size and number of available bytes for the partition, run facter -p pe_postgresql_info on the node that runs the pe-postgresql service.

To speed the migration and optimize queries, clean up the PE-PosgreSQL database prior to upgrade by applying the pe_databases module to nodes running the pe-postgresql service. For best results, apply the module at least one week prior to upgrade to allow the module's maintenance schedule enough time to clean all databases.

After upgrading, you can optionally remove packages and directories associated with older PostgreSQL versions with the command puppet infrastructure run remove_old_postgresql_versions. If applicable, the installer prompts you to complete this cleanup.

Note: If you use an external PostgreSQL instance not managed by PE, you must separately upgrade the instance to PostgreSQL 11. For more information, see Upgrading PostgreSQL.

TLSv1 and v1.1 disabled in PE 2019.1

As of PE 2019.1, TLSv1 and TLSv1.1 is now disabled by default to comply with security regulations. You must enable TLSv1 to upgrade agents on these platforms:

AIX
CentOS 5
RHEL 5
SLES 11
Solaris 10, 11
Windows Server 2008 R2

Certificate architecture and handling in PE 2019.0

PE 2019.0 and later, courtesy of Puppet Server, uses an intermediate certificate authority architecture by default. When you upgrade to PE 2019.0 or later, you have several options for whether and when to adopt the new intermediate certificate architecture:

To upgrade to 2019.0 or later and keep your existing CA, upgrade infrastructure nodes and agents as normal. You can continue to use pre-6.x agents with a Puppet 6.x or PE 2019.0 or later primary server as long as you don't regenerate certificates.
To migrate to 2019.0 or later and keep your existing CA, install the new version and copy /etc/puppetlabs/puppet/ssl from your old primary server. You can continue to use pre-6.x agents with a Puppet 6.x or PE 2019.0 or later primary server as long as you don't regenerate certificates.
To adopt the new CA architecture, upgrade both your primary server and agents, and then regenerate certificates. If you don't upgrade all of your nodes to 6.x, don't regenerate certificates, because pre-6.x agents won't work with the new CA architecture.

MCollective removal in PE 2019.0

If you're upgrading from a 2018.1 installation with MCollective enabled, you must take additional steps to ensure a successful upgrade.

Before upgrade

Remove MCollective from nodes in your infrastructure. If any nodes are configured with MCollective or ActiveMQ profiles when you attempt to upgrade, the installer halts and prompts you to remove the profiles. For example, remove PE MCollective node group and any of the deprecated parameters:
- mcollective_middleware_hosts
- mcollective
- mcollective_middleware_port
- mcollective_middleware_user
- mcollective_middleware_password
Tip: If your PuppetDB includes outdated catalogs for nodes that aren't currently being managed, the installer might report that MCollective is active on those nodes. You can deactivate the nodes with puppet node deactivate or use Puppet to update the records.

After upgrade

Manually remove these node groups:
- PE MCollective
- PE ActiveMQ Broker
- Any custom node group you created for ActiveMQ hubs
If you customized classification with references to MCollective or ActiveMQ profiles, remove the profiles from your classification. In this version of PE, nodes that include MCollective or ActiveMQ profiles trigger a warning during agent runs. Future versions of PE that remove the profiles completely can trigger failures in catalog compilation if you leave the profiles in place.

Removing MCollective

Remove MCollective and its related files from the nodes in your infrastructure. You must have PE version 2018.1.1 or later to complete this task.

Note: This procedure does not remove your MCollective or ActiveMQ log files.

In the console, click Node groups, and select the node group PE Infrastructure.

On the Configuration tab, find the puppet_enterprise class. Select the mcollective parameter and edit its value to absent.


Class	Parameter	Value
`puppet_enterprise`	`mcollective`	`absent`

Click Add parameter and commit the change.
Set up a job and run Puppet on the PE Infrastructure node group to enforce your changes.

Results

The server components of MCollective, including pe-activemq and the peadmin user, are removed from the primary server and the MCollective service on agents is stopped. You must complete the upgrade to 2019.0 or later to completely remove MCollective from agents.

Test modules before upgrade

To ensure that your modules work with the newest version of PE, update and test them with Puppet Development Kit (PDK) before upgrading.

Before you begin

If you are already using PDK, your modules should pass validation and unit tests with your currently installed version of PDK.

Update PDK with each new release to ensure compatability with new versions of PE.

Download and install PDK. If you already have PDK installed, this updates PDK to its latest version. For detailed instructions and download links, see the installing instructions.
If you have not previously used PDK with your modules, convert them to a PDK compatible format. This makes changes to your module to enable validation and unit testing with PDK. For important usage details, see the converting modules documentation.
For example, from within the module directory, run:
```
pdk convert
```
If your modules are already compatible with PDK, update them to the latest module template. If you converted modules in step 2, you do not need to update the template. To learn more about updating, see the updating module templates documentation.
For example, from within the module directory, run:
```
pdk update
```
Validate and run unit tests for each module, specifying the version of PE you are upgrading to. When specifying a PE version, be sure to specify at least the year and the release number, such as 2018.1. For information about module validations and testing, see the validating and testing modules documentation.
For example, from within the module directory, run:
```
pdk validate
pdk test unit
```
The pdk test unit command verifies that testing dependencies and directories are present and runs the unit tests that you write. It does not create unit tests for your module.
If your module fails validation or unit tests, make any necessary changes to your code.

Results

After you've verified that your modules work with the new PE version, you can continue with your upgrade.

Upgrade PE

Upgrade PE infrastructure components to get the latest features and fixes. Follow the upgrade instructions for your installation type to ensure you upgrade components in the correct order.

Before you begin

Review the upgrade cautions for major changes to architecture and infrastructure components which might affect your upgrade.

Upgrade a standard installation

To upgrade a standard installation, run the PE installer on your primary server, and then upgrade any additional components.

Before you begin

Back up your PE installation.

If you're upgrading a replica, ensure you have a valid admin RBAC token. If you're upgrading from 2018.1, the RBAC token must be generated by a user with Job orchestrator and Node group view permissions.

Remove from the console (in the PE Master node group), Hiera, or pe.conf any agent_version parameters that you've set in the pe_repo class that matches your infrastructure nodes. Doing so ensures that upgrade isn't blocked by attempting to download a non-default agent version for your infrastructure OS and architecture.

When you upgrade PE, previous versions remain at /opt/puppet/packages/public. To save space, you can remove files that are no longer needed from this directory.

Optional: Speed upgrade by cleaning up PuppetDB reports. On your primary server, run /opt/puppetlabs/bin/puppetdb delete-reports
If the command fails to execute, you're likely using a version of PuppetDB that doesn't yet include the command. See Upgrade cautions for manual steps.
Download the tarball appropriate to your operating system and architecture.
Unpack the installation tarball: tar -xf <tarball>
You need about 1 GB of space to untar the installer.
From the installer directory on your primary server, run the installer and follow the CLI instructions to complete your server upgrade:
```
sudo ./puppet-enterprise-installer
```
- If you want to specify a different pe.conf file other than the existing file, use the -c flag: sudo ./puppet-enterprise-installer -c <FULL PATH TO pe.conf>
  With this flag, your previous pe.conf is backed up to /etc/puppetlabs/enterprise/conf.d/<TIMESTAMP>.conf and a new pe.conf is created at /etc/puppetlabs/enterprise/conf.d/pe.conf.
Upgrade these additional PE infrastructure components.
- Agents
- PE client tools — On unmanaged nodes only, re-install the version of client tools that matches the PE version you upgraded to. Client tools are automatically updated on infrastructure nodes and managed nodes when you upgrade PE.
In disaster recovery installations, upgrade your replica.

The replica is temporarily unavailable to serve as backup during this step, so time upgrading your replica to minimize risk.
1. On your primary server logged in as root, run: puppet infrastructure upgrade replica <REPLICA_FQDN>
  To specify the location of an authentication token other than the default: puppet infrastructure upgrade replica <REPLICA_FQDN> --token-file <PATH_TO_TOKEN>
2. After the replica upgrade successfully completes, verify that primary and replica services are operational. On your primary server, run: /opt/puppetlabs/bin/puppet-infra status
3. If your replica reports errors, reinitialize the replica. On your replica, run: /opt/puppetlabs/bin/puppet-infra reinitialize replica -y

Upgrade a large installation

To upgrade a large installation, run the PE installer on your primary server, and then upgrade compilers and any additional components.

Before you begin

Back up your PE installation.

Ensure you have a valid admin RBAC token in order to upgrade compilers or a replica. If you're upgrading from 2018.1, the RBAC token must be generated by a user with Job orchestrator and Node group view permissions.

When you upgrade PE, previous versions remain at /opt/puppet/packages/public. To save space, you can remove files that are no longer needed from this directory.

Optional: Speed upgrade by cleaning up PuppetDB reports. On your primary server, run /opt/puppetlabs/bin/puppetdb delete-reports
If the command fails to execute, you're likely using a version of PuppetDB that doesn't yet include the command. See Upgrade cautions for manual steps.
Download the tarball appropriate to your operating system and architecture.
Unpack the installation tarball: tar -xf <tarball>
You need about 1 GB of space to untar the installer.
From the installer directory on your primary server, run the installer and follow the CLI instructions to complete your server upgrade:
```
sudo ./puppet-enterprise-installer
```
- If you want to specify a different pe.conf file other than the existing file, use the -c flag: sudo ./puppet-enterprise-installer -c <FULL PATH TO pe.conf>
  With this flag, your previous pe.conf is backed up to /etc/puppetlabs/enterprise/conf.d/<TIMESTAMP>.conf and a new pe.conf is created at /etc/puppetlabs/enterprise/conf.d/pe.conf.
To upgrade compilers, on your primary server logged in as root, run: puppet infrastructure upgrade compiler <COMPILER_FQDN-1>,<COMPILER_FQDN-2>
- To upgrade all compilers simultaneously: puppet infrastructure upgrade compiler --all
- To specify the location of an authentication token other than the default: puppet infrastructure upgrade compiler <COMPILER_FQDN> --token-file <PATH_TO_TOKEN>
Upgrade these additional PE infrastructure components.
- Agents
- PE client tools — On unmanaged nodes only, re-install the version of client tools that matches the PE version you upgraded to. Client tools are automatically updated on infrastructure nodes and managed nodes when you upgrade PE.
In disaster recovery installations, upgrade your replica.

The replica is temporarily unavailable to serve as backup during this step, so time upgrading your replica to minimize risk.
1. On your primary server logged in as root, run: puppet infrastructure upgrade replica <REPLICA_FQDN>
  To specify the location of an authentication token other than the default: puppet infrastructure upgrade replica <REPLICA_FQDN> --token-file <PATH_TO_TOKEN>
2. After the replica upgrade successfully completes, verify that primary and replica services are operational. On your primary server, run: /opt/puppetlabs/bin/puppet-infra status
3. If your replica reports errors, reinitialize the replica. On your replica, run: /opt/puppetlabs/bin/puppet-infra reinitialize replica -y

What to do next

Optionally convert legacy compilers to the new style compiler running the PuppetDB service.

Upgrade an extra-large installation

To upgrade an extra-large installation, run the PE installer first on your PE-PostgreSQL node, then on your primary server, and then upgrade any additional components.

Before you begin

Back up your PE installation.

Ensure you have a valid admin RBAC token in order to upgrade compilers. If you're upgrading from 2018.1, the RBAC token must be generated by a user with Job orchestrator and Node group view permissions.

When you upgrade PE, previous versions remain at /opt/puppet/packages/public. To save space, you can remove files that are no longer needed from this directory.

Optional: Speed upgrade by cleaning up PuppetDB reports. On your primary server, run /opt/puppetlabs/bin/puppetdb delete-reports
If the command fails to execute, you're likely using a version of PuppetDB that doesn't yet include the command. See Upgrade cautions for manual steps.
Download the tarball appropriate to your operating system and architecture.
Unpack the installation tarball: tar -xf <tarball>
You need about 1 GB of space to untar the installer.
Upgrade your PostgreSQL node.
1. Ensure that the pe.conf file on your PostgreSQL node is up to date by running puppet infrastructure recover_configuration on your primary server, and then copying /etc/puppetlabs/enterprise/conf.d to the PostgreSQL node.
2. Copy the installation tarball to the PostgreSQL node, and from the installer directory, run the installer: sudo ./puppet-enterprise-installer
From the installer directory on your primary server, run the installer and follow the CLI instructions to complete your server upgrade:
```
sudo ./puppet-enterprise-installer
```
- If you want to specify a different pe.conf file other than the existing file, use the -c flag: sudo ./puppet-enterprise-installer -c <FULL PATH TO pe.conf>
  With this flag, your previous pe.conf is backed up to /etc/puppetlabs/enterprise/conf.d/<TIMESTAMP>.conf and a new pe.conf is created at /etc/puppetlabs/enterprise/conf.d/pe.conf.
To upgrade compilers, on your primary server logged in as root, run: puppet infrastructure upgrade compiler <COMPILER_FQDN-1>,<COMPILER_FQDN-2>
- To upgrade all compilers simultaneously: puppet infrastructure upgrade compiler --all
- To specify the location of an authentication token other than the default: puppet infrastructure upgrade compiler <COMPILER_FQDN> --token-file <PATH_TO_TOKEN>
Upgrade these additional PE infrastructure components.
- Agents
- PE client tools — On unmanaged nodes only, re-install the version of client tools that matches the PE version you upgraded to. Client tools are automatically updated on infrastructure nodes and managed nodes when you upgrade PE.

What to do next

Optionally convert legacy compilers to the new style compiler running the PuppetDB service.

Upgrading PostgreSQL

If you use a PostgreSQL instance not managed by PE, you must take extra steps to upgrade PostgreSQL.

Note: If you use the default PE-PostgreSQL database installed alongside PuppetDB, you don't have to take special steps to upgrade PostgreSQL. If you use a standalone PE-PostgreSQL node, follow the steps to upgrade an extra-large installation.

You must upgrade a PostgreSQL instance not managed by PE only when there's an upgrade to PostgreSQL in PE. To upgrade an unmanaged PostgreSQL instance, use one of these methods:

Back up databases, wipe your old PostgreSQL installation, install the latest version of PostgreSQL, and restore the databases.
Back up databases, set up a new node with the latest version of PostgreSQL, restore databases to the new node, and reconfigure PE to point to the new database_host.
Run pg_upgrade to get from the older PostgreSQL version to the latest version.