Neon statusManual checkVerified May 24, 2026Source: neonstatus.com
Is Neon down?
Live status check · No widespread incident reported.
Most recent manual verification mirrored from neonstatus.com. Below: the components that can break, what to do during an incident, and how to keep your scheduled jobs running independently.
About Neon status
Neon is serverless Postgres with separate compute and storage. Incidents are usually scoped to one region or one component (compute autoscale, storage, the control plane); the database is durable across compute restarts, so most outages affect availability rather than data.
Components and surfaces that can break
- Surface 01
Database endpoints (postgres connections).
- Surface 02
Compute autoscale, branching, storage layer.
- Surface 03
Console, the Neon CLI, the GitHub integration.
What to do if Neon is actually down
Confirm your project's region
Neon publishes status per region (US East, EU Central, etc.). If your project is in eu-central-1 and the status page lists a us-east-2 incident, the issue is elsewhere — probably your connection string, pooler config, or app code.
Watch for cold-start spikes on recovery
Neon's compute scales to zero by default. After a regional incident, every project resumes at once. Expect connection latency to be higher than usual for the first few minutes while everyone wakes up.
Use the pooler endpoint, not the direct one
If you've been using the direct connection string from a serverless function, switch to -pooler. Pooled connections handle Neon restarts and brief blips far better than direct ones.
Keep firing through the incident
Schedule your own Neon health check with Crontap
Daily database maintenance (VACUUM tuning, partition rotation, archive jobs) is pure HTTP cron when wrapped in a small endpoint. Crontap fires the endpoint, retries on 5xx, and surfaces the SQL error in the alert body so you don't have to dig through Neon's logs for context.
External HTTP cron hits your endpoint on the cadence you pick, stores every response, and emails you the moment a run fails. Pro schedules down to 1 minute; $3.25/mo annual flat for unlimited jobs.
Go deeper
Neon status: FAQ
- Where do I check Neon status?
- neonstatus.com publishes component status per region. The history page tracks past incidents with timestamps and post-mortems.
- Why is my first query slow after a Neon outage?
- Neon compute scales to zero when idle. After an incident clears, your compute has to resume before serving the first query — typically a 1-2 second wait. Connection poolers and 'keep warm' probes both mitigate it.
- Can I monitor a Neon database from outside Neon?
- Yes. Set up a Crontap probe against a /api/db-health endpoint on your app that runs SELECT 1; on the Neon connection. Run it every minute; you'll catch connectivity issues, pool exhaustion, and storage degradation before they cause customer impact.