The Puppet Application Manager (PAM) UI provides administration functionality where you can access and manage your Puppet applications.

PAM console menu

Use the console menu at the top of the Puppet Application Manager UI to manage Puppet Application Manager itself. It has three tabs of interest to us:

  • Use the Dashboard tab to:
    • Manage your applications
    • See version history
    • Set application configuration settings
    • Access support bundles for troubleshooting
    • Manage licenses
    • View files
    • Configure registry settings
  • Use the Cluster Management tab to view current information on the nodes in your cluster. You can also use this tab to drain, and add nodes to your cluster.
  • Use the Snapshots tab to create point-in-time backups of your deployment, which can be used to roll back to a previous state, or restore your installation into a new cluster for disaster recovery. For more information, see Backing up PAM using snapshots.

You can also use the console menu to Add a new application and to log out of Puppet Application Manager.

Application monitoring graphs

When you have Prometheus installed, the Dashboard tab has an Application sub-tab that provides several simplified graphs for tracking overall health of the system.

  • Node CPU Usage (%) shows when hosts are getting overwhelmed (high % usage).
  • Node Memory Usage (%) shows when hosts are reaching full memory capacity that may result in processes being killed due to out-of-memory errors.
  • Node Available Storage (%) shows when hosts are running out of storage. At 15%, pods may start to be evicted or reads/writes on databases are paused until more storage is made available.
  • Volume Available Storage (%) shows when application persistent volumes are getting full (low %) that may lead to problems with a particular application. Note that -
    Note: As of the 30 June 2021 Puppet Application Manager release, the monitoring/Prometheus-Kubernetes pods limit their storage use and are expected to never fall below 10% available storage.

Puppet Application Manager HA architectures include Prometheus and Grafana. Metrics about how the system is working are sent to Prometheus, and can be displayed with Grafana. Grafana credentials are printed during install, or can be retrieved later with the following command:

kubectl -n monitoring get secret grafana-admin -o go-template='{{index .data "admin-user"|base64decode}}:{{index .data "admin-password"|base64decode}}'