Multi agents and deployment

[subhojit777]

We are assessing langgraph for a use case involving agentic concept. We will self host the agents. I was going through https://langchain-ai.github.io/langgraph/cloud/deployment/setup/ and I had some questions.

1. Multiple agent in langgraph.json
   I noticed multiple agent ids can be defined in a single langgraph.json https://langchain-ai.github.io/langgraph/cloud/reference/cli/#configuration-file. I could not find how the agents will communicate with each other. Or, are they supposed to work independently? They are only kept separate so that their deployments can be managed separately based on dependencies.

2. How to self-host?
   Do I only need to run langgraph up and use the endpoint for communicating with agent? This link is no longer working https://langchain-ai.github.io/langgraph/cloud/deployment/self_hosted/. Can I get the steps please.

3. Multiple agent deployed in separate server?
   Would it be advisable to deploy each agent in separate servers? based on traffic, etc. If so, how it can be achieved?

Thank you.