With Puppet officially announcing the May 8th general release of Puppet Discovery, I wanted to take a moment to share my experience with their publicly available tech preview.
The goal of my project was to help a customer identify what services were running on their infrastructure and any packages which would need updating along with any additional facts I could find. I had several options at my disposal, but the common challenge among them was to present a human readable and searchable format where my customers could see in real time what was running in their infrastructure. Before beginning to code a database populated by ruby scripts, I decided to give Puppet Discovery a try against some test VMs.
From installation to deployment, the process was relatively quick. Within minutes of installing the virtual box on my MacBook Pro, I was adding hosts in CSV format along with my private RSA key. I had options to add WinRM credentials, RSA private key, or AWS account information. Almost immediately, server data started to populate my dashboard and I could see the details I was seeking as I navigated the pages.
The uncluttered dashboard layout was intuitive with details a click away. Selecting a package would bring me to a list of all the servers with that package installed. I could now select the sub group of Web Servers or group by OS type.
In the tech preview I had limited OS versions available. After an email to Puppet I had some answers about upcoming support and features that are planned for the release version. More granular OS details will be available on the server lists in the general release version and I was told Juniper Switch support is planned for after initial product launch. Puppet Discovery can already give real-time stop/start/restart on any service discovered. It is planned for Puppet Discovery to work in conjunction with Puppet Enterprise and support tasks from Puppet Bolt. In the tech preview, the option to deploy the puppet agent is available already to demonstration additional functionality.
I would encourage any DevOps Engineer working with large scale automated infrastructure both on premise and AWS, to look at what is possible with Puppet Discovery.