doesn't take much time to learn, my junior teammates use the PE Orchestrator without problems. it's great for running jobs on puppet-managed machines.

for some reason one of the most useful modules doesn't come bundled with PE, https://forge.puppet.com/modules/puppetlabs/exec/ that is, "just run a command"

"being quick and easy to learn" doesn't change "not something we need" ... and i don't see how

puppet task run exec command="df -h" --nodes xyz

is any better than

ssh xyz df -h

it's a few clicks in a web interface, but your loss 🤷

no worries, i appreciate the suggestions ... right now my biggest problem is that i don't have time to learn about all the new stuff

I think that’s the disconnect. Many of these tools were releasse years upon years ago. Most people no longer view them as “new”

Orchestrator first shipped about a decade ago in PE 2015.3 🙂

and choria is a descendant of mco

Tasks and plans run over pcp and the pxp-agent, increasing security and reliability by not requiring individual user keys all over. They are auditable because they're stored in the PE logs. They can be run in parallel across many systems. The results are shareable through the interface. For juniors, they can be guard railed by only allowing specific flags and parameters. They're actually quite useful from an organizational standpoint, even if not specifically from a long time administrator who knows how to do everything by hand standpoint.

I am the only one who has stumbled on the install script for Puppet Enterprise 2025 failing due to a dependency on chkconfig? Apparently during an installation of pe-postgres. Time to sync up with the 21th century perhaps? 🙂

My agent server is looking at a branch created in Azure which was removed after a Pull request was completed. However after deleting the branch, I was expecting the server to look at production branch when running puppet agent -t but it still appears to be looking at an old branch which has been removed. I tried to research how to fix this myself but cant figure out a way around. Can someone please advise what can be done to fix this issue?

you’ll likely need to delete that branch in your control_repo as well, run a puppet-code deploy on the server, and then the branch should stop being known about by the Puppet server

yeah tried this and usually it picks up the production branch but its still somehow trying to pickup the deleted branch

I know open source has r10k artifacts stashed in /opt/puppetlabs/puppet/cache/r10k. Maybe somnething similar for Code Manager?

Assuming some previous fetch was corrupted or something I ran rm -rf /opt/puppetlabs/server/data/code-manager/git/*

Then when i try to run puppet-code deploy --all --wait it has errors

the environment might be set to the non-production one in the PE console for the node

image.png

This doesnt pickup production which is strange

We do have that branch in azure and tests show it can ping it etc

agent-specified

means the node classifier, i.e. the Console, will use the environment that the agent requests. So whatever is in the agent's

puppet.conf

is what will be used for the environment.

puppet.conf does not have any code to explicitly pickup production

Usually it just look at production

The list should show production in its values like it did before but it does not so trying to fix it

one thing to be aware of… the console has a cache as that can be an expensive request if there’s many environments, on the classes tab of the node group detail screen there’s a refresh button

Screenshot 2025-11-13 at 14.37.56.png

I do know but that doesnt fix the issue of production not showing

to narrow things down, what does puppetserver list?

Untitled

did you try running explicitly with production, like

puppet agent --test --environment production

and then see if the next runs without explicit

environment

setting keep using production