Contact the Anaconda implementation team before you begin for assistance upgrading your version of Data Science & AI Workbench. Follow along with these instructions as an Anaconda implementation team member guides you through the upgrade process.
Current versions of Workbench no longer support Postgres versions 9.6 and older. To align with these changes, you must update your Postgres to a supported version. Verify your Postgres version is compatible before you begin your upgrade.

​
Prerequisites

  • You must have jq installed.
  • You must have a service account with sudo access on your Kubernetes (K8s) master node.
  • Your K8s master node must be configured with a DNS A record. The A record sets the domain name you use for your Workbench instance.
  • If you are upgrading Workbench on a system with multiple nodes, you must verify the clock on each node is in sync with the others. Anaconda recommends using Network Time Protocol (NTP) to synchronize computer system clocks automatically over your network. For more information on installing and using Chrony to manage the NTP, see the instructions provided here.
  • You must have the ae-preflight package installed. For more information, see Workbench pre-flight check.
  • Create a backup of your anaconda-enterprise-anaconda-platform.yml ConfigMap by running the following command:
kubectl get cm anaconda-enterprise-anaconda-platform.yml -o json | jq -r '.data["anaconda-platform.yml"]' > configmap-backup.yml

​
Upgrading

After you have verified that your system meets all of the installation requirements, you’re ready to upgrade the cluster. There are two basic types of upgrades for Gravity users: in-place upgrades and fresh reinstallation upgrades. Follow along with the instructions here as your Anaconda implementation team member guides you through upgrading your software.
  • Project sessions are terminated during the upgrade process! Because of this, it is important to stop all sessions prior to upgrading. If you do not, sessions that are terminated as part of the upgrade process must be restarted manually post upgrade.
  • Sessions can be stopped programmatically using ae5-tools by running the following command in a terminal that has access your Workbench cluster over the network:
ae5 session list --columns=id --no-header | xargs -n1 ae5 session stop --yes
In-place upgrades of Workbench are not supported for versions that are moving from Gravity 6 to Gravity 7. Instead, you must perform a fresh reinstallation to upgrade your software. If you do not, your installation will break. Check the version of Gravity and Workbench before you begin, and choose the upgrade process that best suits your needs.In-place upgrades are also not supported when upgrading from Workbench 5.6.x to 5.6.2. Please perform a fresh reinstallation to upgrade between these versions.
In-place upgrades are performed while the software is still running. To perform an in-place upgrade:
  1. Create a backup of your current instance of Workbench.
  2. Log in to a service account with sudo access on the master node running your Workbench software.
  3. Download the installer file by running the following command: 
    # Replace <INSTALLER_LOCATION> with the provided location of the installer file
curl -O <INSTALLER_LOCATION>
  4. Decompress the installer file by running the following command: 
    # Replace <INSTALLER> with the installer file you just downloaded
tar xvzf <INSTALLER>
  5. Enter the installer directory by running the following command: 
    # Replace <INSTALLER> with the version of your installer
cd <INSTALLER>
  6. Run the following command to verify your environment is properly prepared: 
    ae-preflight
    If the check returns an overall result of WARN, you can view the results of the check by running the following command: 
    cat results.txt
    If necessary, make applicable corrections to properly prepare your environment to meet the installation requirements. Once you’ve verified that your environment is properly configured, you can begin the upgrade process.
  7. To start the upload and upgrade process, run the following commands: 
    sudo ./upload
sudo ./gravity upgrade
    The upgrade process can take up to an hour or more to complete, primarily due to the upload step. You can view the status of the upgrade process at any time by running the following command: 
    sudo watch ./gravity plan
    Once the upgrade process is complete, the pods begin to initialize on their own, but this process takes some time to finish. Monitor the pods’ status by running the following command: 
    sudo watch kubectl get pods
    If you encounter errors while doing your in-place upgrade, you can view which phase of the upgrade failed by running the following command: 
    sudo ./gravity plan
    You can return to any phase of the upgrade process by running the rollback command against the name of the phase as it’s listed in the Phase column of the ./gravity plan commands’ return: 
    # Replace <NAME_OF_PHASE> with the name listed in the Phase column
sudo ./gravity plan rollback --phase=/<NAME_OF_PHASE>
    After addressing any errors, resume the upgrade by running the following command: 
    sudo ./gravity upgrade --resume --force
    Once you have resolved your errors, or if no errors have occurred, it’s time to verify your installation.

​
Verifying your upgrade

  1. Verify all pods are running by running the following command: 
    sudo kubectl get pods
  2. Open a web browser and navigate to your Authentication Center. 
    # Replace <FQDN> with the fully qualified domain name of your Workbench server
https://<FQDN>/auth/admin
  3. Select Users from the Manage menu, then click View all users and verify your users’ data is present.
  4. Open a web browser and navigate to your Workbench URL. Log in using the same credentials you used for your previous installation.
  5. Review the Projects list to verify that all project data has been restored.
  6. Verify your deployments have returned to a started state.

​
Additional configurations

​
TLS/SSL certificates

If you did not configure SSL certificates as part of the post-install configuration, do so now. For more information, see Updating TLS/SSL certificates.

​
External version control repository

Workbench uses configurable parameters in the git: section of the anaconda-enterprise-anaconda-platform.yml configmap to connect to external version control repositories. Verify your parameters are mapped correctly, as described here.

​
Spark/Hadoop

After verifying your installation, run the following command on the master node of the Workbench server: 
# Replace <PATH_TO_SECRETS.yaml> with the path to your anaconda secrets .yaml file
sudo kubectl replace -f <PATH_TO_SECRETS.yaml>
To verify that your configuration upgraded correctly:
  1. Log in to Workbench.
  2. If your configuration uses Kerberos authentication, open a Hadoop terminal and authenticate yourself through Kerberos using the same credentials you used previously. For example, kinit <USERNAME>.
  3. Open a Jupyter Notebook that uses Sparkmagic and verify that it behaves as expected by running the sc command to connect to Sparkmagic and start Spark.

​
Cleaning up

As part of the upgrade process, the script you run automatically removes the unused packages and images from the previous installation and repopulates the registry to include only those images required by the current installation. This helps prevent the cluster from running out of disk space on the master node.