Good morning! I am currently trying to see if there is a way to add a VM to an existing deployment through any means (action, workflow, etc.). I haven't seen anything in the documentation and blogs that I have come across, so if there are any other avenues, I would appreciate being pointed in the right direction!
A little new to vRA. Forgive me if I am missing a large piece of knowledge base.
If you have a blueprint which allows, say, 5 vSphere machines and you only deploy one. You end up (if successful) with one deployment containing one vSphere machine. You can then choose to scale that deployment if you wish to increase the machine count.
By scaling out, you select how many more instances you want.
But all of this is built-in at the blueprint level. When I said "arbitrary" I meant you can't have an existing deployment and then choose to deploy another catalog item and have it join a preexisting deployment from a separate blueprint.
You can if you scale-out an existing deployment. But you can't just select any arbitrarily-deployed VM or VMs and join them to an existing deployment.
Hi, daphnissov:
Thank you for your reply! I looked into scaling out in vRA (i.e. Blogs like https://theithollow.com/2016/10/06/scaling-vrealize-automation/), but I am still a little lost.
When you say I cannot join an arbitrarily deployed-VM to a deployment, what restrictions are there? Must the VM already exist, or must it be newly created? Thank you for your patience!
If you have a blueprint which allows, say, 5 vSphere machines and you only deploy one. You end up (if successful) with one deployment containing one vSphere machine. You can then choose to scale that deployment if you wish to increase the machine count.
By scaling out, you select how many more instances you want.
But all of this is built-in at the blueprint level. When I said "arbitrary" I meant you can't have an existing deployment and then choose to deploy another catalog item and have it join a preexisting deployment from a separate blueprint.
Good afternoon, daphnissov:
Thank you for answering my question about deployments a few days ago. I am still learning about VMware, vRA, and all the likes and am still fairly lost.
Could you confirm if I understand this correctly: Scaling out is the process of adding another VM to a deployment. It doesn't necessarily have to be a clone of the existing VM.
I guess I'm not too clear on what scaling out means. I know you can scale out applications (i.e. Server A cannot handle the load of calls, so you can scale out by adding Server B). Would adding a different server to perform something entirely (i.e. a bonus search server) different still count as scaling out?
I have this situation where there is a blueprint with one template to clone (and the capability to deploy up to 20 VMs). Users can create one VM, but may choose to create another VM on a later date. Is this scaling out?
It may actually help you to read some of the documentation on scaling with in vRA and, to some other extent, scaling as a general concept. When you scale within vRA, it isn't a clone of the system that's already running out there but of whatever the blueprint defined. So, for example, if you have a system of web front ends that are getting their identity through Puppet or even a software component, when you scale a deployment that already contains N number of web frontends, you create additional web frontends with the same identity and in the same LB pool (if that applies to you). This is horizontal scale where you're adding instances. Contrast that with vertical scale where you add more resources to *existing* instances.