Supabase Capacity Outage: What Broke and Who's Affected
July 4, 2026
Supabase has been fighting a capacity problem across nearly every region since June 30. If you tried to create, resize, or restart a project this week, you probably ran into it.
What Broke
Starting June 30, Supabase saw widespread capacity constraints hitting project creation, resizing, restarts, branch provisioning, and restores — across most regions, not just one. Users saw repeated failures on these operations for days.
By July 1–2, capacity started coming back online region by region, but ap-northeast-2, ap-south-1, and eu-north-1 lagged behind the rest.
The Part That's Still Broken
As of July 3, the remaining problem is narrower but nastier: a subset of projects — specifically Medium instances running PostgreSQL earlier than version 17.6.1.121 — are stuck down due to what Supabase calls "infrastructure compatibility constraints." New projects deploy fine. Existing ones on this profile that get restarted or resized may fail.
If your project hasn't been restarted or resized during the incident window, Supabase says it's not affected. If it has, and it's on an older Postgres version on a Medium instance, that's your risk zone.
What to Do Right Now
Check your Postgres version if you're on a Medium instance. If it's below 17.6.1.121, hold off on restarts or resizes until Supabase confirms the fix is fully rolled out — triggering one right now is the thing most likely to land you in the stuck-project bucket.
If you're already stuck, Supabase says its team is actively working through the backlog project by project. Contact support rather than repeatedly retrying — retries didn't help other users in the same spot.
Why This Keeps Happening
Backend-as-a-service platforms concentrate a lot of infrastructure risk in one place. When it's your database, your auth, and your storage all running through the same provider, a capacity problem doesn't just slow you down — it can block you from operating on your own project entirely.
This is the same lesson from the AWS us-east-1 outage in May — worth reading if you're building anything on managed infrastructure without a fallback plan for the provider having a bad week.
Source: Supabase Status
Frequently Asked Questions
What happened with the Supabase outage?
Starting June 30, 2026, Supabase experienced widespread capacity constraints across most regions, causing failures in project creation, resizing, restarts, branch provisioning, and restores.
Is my Supabase project still affected?
As of July 3, 2026, Medium instances running PostgreSQL versions earlier than 17.6.1.121 remain at risk if they are restarted or resized. Existing projects that haven't undergone a restart or resize during the incident are not affected.
How do I avoid the Supabase capacity issue?
Check your Postgres version if you're on a Medium instance. If it's below 17.6.1.121, avoid restarting or resizing until Supabase confirms the fix has fully rolled out, since that action is what's triggering the stuck-project state.
What should I do if my Supabase project is stuck down?
Contact Supabase support instead of repeatedly retrying. Supabase says its team is working through the backlog of affected projects individually, and retries haven't resolved the issue for other users in the same situation.