Saif Ali Shaik —
I've tried deploying Archestra on render.com, but it turns out I cannot expose both :3000 or :9000 ports for admin UI and API access simultaneously. Is there a work around to this?
#general
I watched the stream recording for Archestra, and it was actually great. By the way, what projects are you all working on?
I am using archestra ai's MCP and Agent in my project. So in the submission page it's asking for deployment project link.
Locally i can deploy the agent and MCP servers through archestra development method. So my question is how can i use MCP server and agents through archestra after i deploy my project online? As archestra is not hosted for public that i can use it's api
I've tried deploying Archestra on render.com, but it turns out I cannot expose both :3000 or :9000 ports for admin UI and API access simultaneously. Is there a work around to this?
Is there a way (a API) that where I can get the current logged in user details? such as email address? or any identifier?
Hello @user@user
I’m trying to run Docker quickstart, but it fails with the error “Unable to create a Kind cluster.”
I’ve already enabled Kubernetes in Docker Desktop and restarted it.
The issue still persists even after multiple attempts since yesterday..
👋 Hi everyone! where is the problem statement for the hackathon?
Hello. I am trying to follow the example in the quickstart
But I get errors when I try to install playwright mcp
These are the docker logs:
👀2
7 repliesFollowing the documentation, a basic
/health api endpoint works. But if try to GET /api/organization I see:I'm using the API key at: Settings > Your Account > API Keys > Create API Key
cURL-H 'Authorization: Bearer archestraredactedAMHqcbsPG'```
Is there something I'm missing?
Can I create a user in Archestra programmatically? I was assuming making a POST call to /api/teams/{id}/members woudl also create members
Is it even possible
Team, reading the channel and can't be more excited to see what you're building. 🚀 🚀 🚀
❤️6☺️1
this error is usually caused by insufficient Docker memory/resources on Windows.
but i tried in Ubuntu Wsl, still got the error
try using github codespaces, that might work.
I have ubuntu
Read-only live mirror of Archestra.AI Slack
👋Join the discussion withAI enthusiasts!I've tried deploying Archestra on render.com, but it turns out I cannot expose both :3000 or :9000 ports for admin UI and API access simultaneously. Is there a work around to this?
5 replies
Yup... I also mentioned same issue
so I haven't tried this fully myself yet, but in practice this 👇 should work:
all traffic can go through port 3000. The Next.js frontend already has rewrites configured in
next.config.ts that proxy everything to the backend internally:•
/api/* → backend port 9000•
/v1/* → backend port 9000 (LLM proxy endpoints)•
/health → backend health check•
/ws → WebSocket connectionsSince both the frontend and backend run inside the same container (managed by
supervisord in the Docker image), port 9000 is just an internal port. You only need to expose port 3000 externally, and all API calls will be transparently proxied to the backend within the container.On Render: just set the exposed port to 3000 and make sure
ARCHESTRA_INTERNAL_API_BASE_URL is set to http://localhost:9000 (which should be the default). Everything should work through the single port.tl;dr 🙂
Thanks @user. It turns out the its' easy to exhaust the free plan (512 MB) from render.com to deploy and record a demo. So I think I may have have to go with local version itself. Appreciate you chiming in!
:thumbsup_all:2