I am looking for help as I am new to K8s - I would like to use Pods to house 2 containers - 1 container would be screens that will be available in a UI platform and the 2nd container will be the API calls to the equipment management tool. So let’s say I have a Cisco Viptella and I want to create a UI that my customers will interact with. This customer wants to have Cisco Viptella as an SD-WAN service and in the customer portal, I want screens specific to Viptella. So the Pod would have container 1 (Screens for Viptella) and container 2 (API calls for Viptella controller).
Long story to ask if this is actually possible?
This sounds more like a single container in a pod: a web service that exposes a REST API and they’d interact with your custom web interface via a browser and the RESTful interface via a URI…
Thank you for the response. For additional clarity - so it is possible but I would only need to have a single container for Both the AREST API data and the specific screens representative of the specific element. Basically I have an UI and I want to enable specific screens for a specific product. So screens for Viptella, specific screens for Silverpeak etc… it would be a container for each of the individual elements supporting the specific product REST API’s and screens that will be populated into the top level UI.
I’m not super familiar with viptella, but are you trying to extend the network of the pods to your client networks? If so, theres going to be a lot of complications there, both from a networking and security perspective.
Yeah, I don’t think you’re going to be able to do that without a significant amount of work. Kubernetes is not really intended to mix networks like that. It IS possible, there are a lot of trade offs and considerations to make.
Thank you - this is great info. Sorry to continue to bother you but a little more clarity to my ask - The SD-WAN platforms (A,B,C) are different manufacturers but all deployed across the same telecom network with different orchestrators that are on the telecom providers network. The UI is the customer portal provided by the telecom provider. you can have 1 customer using sdwan platform a and 2nd customer using sdwan platform b. they both access the same customer portal with SSO etc… but their screens for interacting with their sdwan platforms of choice, a or b are specific to their platform. The containers for A and B Sdwan platforms have the API calls that connect to their respective devices and/or orchestrator and have the API call responses feed to an integration platform that then feeds the data to the UI Screens that are built to both appear in the customer portal and interact between the portal and the specific sdwan platform. all on the same management network but logically segmented for customer segmentation using provider vlans. not sure if this makes since or changes your current thoughts. Thank you again.
same managed network == physical network?
It doesn’t really change my thoughts on it, what you want to do IS possible, but you will have to spend a significant amount of time looking into how to integrate k8s into your network topology. Kubernetes is (generally) very flat in its network design, All the pods exist on a network (physical or overlay) where they can talk to all other pods without NAT. Attaching them to specific vlans via a second adapter with something like multus is possible, but it will take a good amount of work plan and execute
