Skip to content

How to scale a A2A production deployment to 100 A2A Server Agents each exposing a different AgentCard? #898

@mescaja

Description

@mescaja

Say i have an quarkus App that implemented A2A SDK with on AgentExecutor and one AgentCard and i would like to scale this deployment to support a higher number of Server Agents.

the one approach that would work would be to deploy one separate Quarkus app per Agent to kubernetes and proxy the connection with a reverse proxy software creating virtual hosts/ingresses to access the Server Agent endpoint (AgentCard and JSONRPC handler)

https://a2a-server-agent-1/.well-known/agent-card.json
https://a2a-server-agent-2/.well-known/agent-card.json
etc

I think that's a lot overhead running a separate quarkus app per agent

is there any best practices in terms of setting up just one single quarkus app supporting multiple AgentCards and JSONRPC Handler endpoints?

is there any reference architecture for a typical A2A production deployment?

Metadata

Metadata

Assignees

No one assigned

    Labels

    No labels
    No labels

    Type

    No type
    No fields configured for issues without a type.

    Projects

    No projects

    Milestone

    No milestone

    Relationships

    None yet

    Development

    No branches or pull requests

    Issue actions