4 out of 5.0, Reviewed Nov 28, 2016
Implementing Puppet for our greenfield of new virtual machines was successful and obstacles were manageable. However, our initial attempt at utilizing Puppet's Roles features was abandoned and we're still working on a second attempt. Our implementation was mostly accomplished internally, but we requested several conference calls which were helpful. After gaining some experience with our greenfield, we then deployed to our brownfield with similar results.
Consider rolling out Puppet to new hosts only at first, preferably in your Development, Testing/QA tiers, in order to refine your package management without interrupting existing services. When your configurations mature, apply to your brownfield.
We have confidence that we can quickly apply security updates across a vast field of servers with many configuration families. Integration with GitHub is especially helpful for day-to-day application-related updates.
The administrative interface can be a little clunky, requiring lots of clicks to drill down to logs related to puppet runs.
I'd appreciate a greater emphasis on logs presentation and reporting in the admin interface -- something more akin to how log aggregation products present data.
If we could start over, we'd do so with a focus on orchestration for applications. That feature set was released soon after we started, and without it we live with configuration-per-VM instead of orchestration by application.
4 out of 5.0, Reviewed Aug 22, 2016
Puppet Enterprise is a great solution that continues to be updated. the company has an excellent vision and strategy.
Make sure you have the correct staff in place for implementation.
It's an industry leading product that is driving change in the organization.
It is still an agent-based product whereas competitors in the market are using SSH to handle creation. They are also focusing on VAST services rather than going deeper on specific technologies. For example Ansible has the ability to manage networks, Puppet Enterprise does not.
Provided less expensive resources for implementation.
We would have waited for 4.0 to be released and then implemented that. We would have also paid a consultant/VAR to do the work and come in on a yearly basis to develop our code pipeline.
Product is capable within the space. It does rely heavily on the community.
Support is good, it's easy to call in and open tickets.