Discovery: the next frontier
Editor's note: We've sunsetted Puppet Discovery and Lumogon is now unmaintained.
We’ve had 12 years of experience at Puppet helping IT professionals cope with the complexity of managing an ever-larger infrastructure and application estate by harnessing the power of automation. In our first dozen years, our core insight was to help our customers define a desired-state model for their infrastructure and applications, and provide an engine to enforce that desired state, prevent configuration drift, accelerate infrastructure and application rollouts, and scale to 100,000 nodes.
Over the course of this journey, we were fortunate to have an incredible open source community that helped us refine our offering and develop nearly 5,000 modules covering every aspect of systems management, so new Puppet users hit the ground running, instead of having to recreate content that already exists.
As we’ve grown our user base, the common questions we hear focus on how to get started with automation and modeling — specifically, “What resources and applications do I have running today?” and “How do I bring these resources under management?” We use the term discovery to refer to these problems.
Puppet’s main focus as a product thus far has been to help users define a desired-state model for their resources, allowing Puppet to ensure those resources are in line with that model. To do this, the Puppet agent gathers tons of operational data from each system on a regular basis, and pushes it up to the Puppet Server.
A first step towards helping our customers with the process of discovery is to show them all the packages they are running across all their systems — whether or not they have modeled those packages. This allows an organization to derive many important insights, including the version distribution of those packages, and even which packages may be contain known vulnerabilities. We’re excited to announce that with the package inspector feature in the 2017.2 Puppet Enterprise release, we have delivered the initial down payment on our vision for discovery.
Discovering cloud resources
Public clouds greatly reduce the barriers to standing up infrastructure for new projects — all you need is a credit card. As organizations increasingly move their application estate to public clouds, IT departments find themselves having to go back and discover all of these new infrastructure resources and applications. That's why we’re excited to announce a limited preview of our Puppet Cloud Discovery™ effort, which helps operations teams get a better sense of what they have running in the cloud. You can learn more about Puppet Cloud Discovery in this blog post.
Containers are quickly emerging as a new way of packaging applications, with the container image being the artifact that is handed off from development into operations. Containers are attractive in no small measure because they boot much faster than virtual machines, and you can achieve an order-of-magnitude increase in density by running containers instead of VMs on your infrastructure.
Containers represent a big opportunity for organizations that want to accelerate their product delivery efforts, but today they are largely black boxes from an operational standpoint. What’s more, since there are so many of them, and they tend to have shorter lifespans than VMs, running containers in production brings unique challenges for an operations team. It’s for these reasons that we're happy to announce the initial release of Lumogon™, a new effort that focuses on discovering what’s running inside all of these ephemeral black boxes, giving operators visibility and control of their container estate.
Puppet and discovery
Discovery is the next frontier for Puppet. Discovery helps IT professionals get a better sense for what they have running — whether it's traditional applications or microservices, whether they're running on-premises or in the cloud, in VMs or containers. But Puppet’s true potential lies in making the transition between discovering resources and modeling them a seamless one. It’s good to find all the packages that may be vulnerable, but it’s even better if we can auto-suggest a module that will help bring those packages up to date and eliminate the vulnerability, all in the same user experience. This is our vision for how Puppet products will evolve over the next few releases.
Puppet 5 and Puppet Enterprise 2017.2
Discovery isn't the only thing we're announcing today. We are also proud of the new Puppet 5 Platform and Puppet Enterprise 2017.2 releases, which give you greater visibility and control of your infrastructure. We'll appreciate your feedback on these, and our other discovery-oriented projects.
Omri Gazitt is the chief product officer at Puppet.