Ansible Automation Platform 2.3 Walkthrough

This video will cover a full walkthrough of the Ansible Automation Platform version 2.3
The presenter is Alex Dworjan
Github: github.com/shadowman-lab
www.ansible.com/blog/ansible-automation-platform-a-video-tour
www.ansible.com/blog/whats-new-in-red-hat-ansible-automation-platform-2.3
www.ansible.com/blog/fun-with-private-automation-hub-part-1
www.ansible.com/blog/fun-with-private-automation-hub-part-2
www.ansible.com/blog/whats-new-in-ansible-automation-platform-2.1-automation-mesh
www.ansible.com/blog/peeling-back-the-layers-and-understanding-automation-mesh

[@cloudguru123]

Awesome walkthrough of 2.3. I have few questions on 2.3 version. 1. How many execution nodes can 2 controllers handle ? 12-15 exe nodes? 2. In case i want to add more controller nodes or more execution nodes? Do i just update the inventory file (incl new exe nodes) and re-run the setup on controller node ? 3. I have single node controller, Can this be scaled to 2 node controller + 15 nodes executions just by re-runing the setup with updated inventory file?

[@alexdworjan]

1. There is no limit, it will depend on the resources allocated to those 2 controller nodes (do they have the capacity to manage the control workload and message routing. 2. Yes, just update the inventory file and re-run the setup 3. You can only scale to a cluster if you currently have a single node controller with an external database. If you have that, you can scale as much as you need assuming the VMs have the appropriate resources

[@cloudguru123]

@@alexdworjan Thank You for your response. 1. I dont have external DB, im using DB internally within the single controller node. 2. Execution nodes are remote or in another DC. Do they have thier own internal postgres DB ? Or there will be no db for execution nodes. I assume DB is only for control / hybrid nodes only.

[@alexdworjan]

@@cloudguru123 There is no supported way to scale from a single node instance to a clustered install. There are ways you can do it, but it will require a backup and restore to a new environment: access.redhat.com/solutions/2989021

[@rasalerajesh]

@@alexdworjan Can we have Hybrid nodes running as active active in 2 different DCs? In case one DC is down, We can still have AAP running ? If this is possible How would the postgres DB replicated between the sites? How can we achieve this ?

[@alexdworjan]

@@rasalerajesh Latency is going to be the biggest issue if you have nodes in separate DC since they all have to communicate back to the DB. If you manage AAP itself as infrastructure as code, you could have two separate clusters with two separate DB in two separate DC but they are configured exactly the same. Then you could place a LB or API gateway in front to direct traffic to whichever cluster is active. You could also set up postgres DB replication, but that's outside of what RH provides (I admit I am not a database specialist so I honestly don't know)