Upgrading Automation Suite

Important:

Upgrading Automation Suite enables maintenance mode on the cluster, which causes downtime for the entire upgrade duration.

If your Automation Suite cluster was ever on 2021.10.3 or 2021.10.4, you must take additional steps to downgrade the Ceph objectstore version before upgrading to a newer Automation Suite version. For instructions, see Downgrading Ceph from 16.2.6 to 15.2.9.

Requirements

• Make sure you meet the hardware and software requirements for new Automation Suite version.

• Download the required installer and bundle files:

  Installation type	Files to download
  Online	sf-installer.zip – Mandatory. See sf-installer.zip for download instructions.
  Offline	sf-installer.zip – Mandatory. See sf-installer.zip for download instructions. sf-infra.tar.gz – Mandatory. See sf-infra.tar.gz for download instructions. sf.tar.gz – Mandatory. See sf.tar.gz for download instructions.

  Important:

  It is strongly recommended to create a backup of the cluster and the SQL database before upgrading Automation Suite. This is to ensure you can restore the cluster if something goes wrong during the upgrade operation.

  Cluster data is backed up on NFS server. Make sure you have configured NFS server before proceeding further.

  Make sure to copy the value of global.userInputs.identity.krb5KeytabSecret to global.kerberosAuthConfig.userKeytab, if you simultaneously meet the following requirements:

  • you configured the Active Directory integration using username and password
  • you have Windows authentication enabled
  • you do not use SQL integrated authentication

Overview

The following illustration shows the steps you need to take for a successful upgrade operation.

For a complete upgrade operation, follow these steps:

1. Configure a backup before upgrading the cluster to a new version, if possible.
2. Enable maintenance mode on the cluster, by bringing down the cluster network, to avoid any writes during upgrade. This step ensures the upgrade is consistent.
3. Upgrade Kubernetes and other infrastructure components on all the cluster nodes.
4. For offline installations, hydrate Private Docker Registry with the new container images.
5. Upgrade the fabric and shared components on the cluster.
6. Upgrade the services (UiPath products).
7. Bring back the cluster by disabling the maintenance mode on the cluster.