Closed mrvdot closed 10 months ago
For a little extra context, I tested this with a brand-new repo / SB install, and same problem. I followed the steps at https://supabase.com/docs/guides/functions/quickstart and got the same error when I got to supabase start
I also tested after switching to Node v21 (latest) and Node v18 using nvm
, and still same error. Didn't expect anything since SB uses docker, but wanted to rule it out anyway
After more digging, this appears to same as these issues:
Will try updating docker
Final note, updating docker to v24 fixed this (Ubuntu defaults to v20, and even though I was using Docker's official APT repo, it was never flagged for upgrade).
For anyone else who hits this, the TLDR is a combination of old Docker (#ThanksUbuntu), a seemingly innocuous update in Node 18, and relying on a base node image rather than pinning to a specific version all combined to cause everything to blow up.
Fix: Update Docker via steps at https://docs.docker.com/engine/install/ubuntu/
Describe the bug Running
supabase start
appears to start, then dies while waiting for health checks, specifically:service not healthy: [supabase_edge_runtime_provider-admin supabase_pg_meta_provider-admin supabase_studio_provider-admin]
To Reproduce Steps to reproduce the behavior:
supabase-cli
in packagesupabase start
(have no functions written and have made no modifications to theconfig.toml
)