Event-driven ansible

@Mark Hensler the Settings v4 roll-out happened without the full coordination of documentation. We are in the process of fixing it in the upcoming sprint.

I don’t know when it dropped, but it looks like documentation is now available:

Event-Driven Ansible Integration Setup | LogicMonitor

Hey @Stuart Weenig 

Event Driven Ansible (EDA) will only be part of Ansible Automation Platform (Redhat supported) and Ansible Galaxy (community supported). 

https://www.redhat.com/en/technologies/management/ansible/ansible-vs-red-hat-ansible-automation-platform

Inside of the LM config you’ll be able to define raw alert data (tokens) or key/value pairs. You’ll be able to check this out in the upcoming Settings beta. 

Cheers

I’ve said it before and I’ll say it again: LM should embrace containerization on the collector. If the collector shipped with the ability to run containers, many problems could be solved:

1. Ansible playbooks could be run without installing ansible. Just run a container that runs playbooks.
2. Wanna run a datasource in any language? Just find the appropriate container and provide your script. Need to run Python? No problem and no need to pre-install anything on the collector. Just select the python container and provide your script. The datasource could even have a drop down that lets you select from images available on the docker hub (or your image repo of choice), type in your own image name (provided you’ve built it on the collector), or provide a Dockerfile to build the image at runtime (in case you need python with the requests library). 

I think a lot of the knowledge around containers at LM is focused around containers that run services that run 24/7. That’s only one use of containers, ephemeral containers is another good use case and would be a great way to instantly expand the supported languages to, well, all of them.

LM has a lot of knowledge around k8s and that could be leveraged to have containerization available on the collector to simply run adhoc containers for specific collector tasks (even alert/insight responses).  

I saw this, today:

When I searched for documentation, I am prompted for a password:

https://www.logicmonitor.com/support/settings-new-ui-overview

Anyone care to share?😋

I would still prefer to see a script (ActionSource?) as a step in an escalation chain.  As an MSP, centralized services like Ansible are more difficult to implement at scale.  We’d need to stand up an Ansible deployment for 150 environments.  Whereas, we’ve already got a collector in place, and all of the code exists in the product to make this a (relatively) simple add for LM.

I am very interested in hearing/seeing more about this.

Hoping you can provide more of an explanation than that.