Disaster recovery
Your data is always safe
Section titled “Your data is always safe”ReflexDB does not store your data persistently. The in-memory snapshot on each running instance is a read-only copy of your MySQL database — it is rebuilt from your source database on each provision or rebuild.
If a ReflexDB instance goes down: your MySQL database is completely unaffected. All your data is still there, exactly as it was.
If the ReflexDB control plane goes down: running instances continue to serve queries independently. The control plane is only involved in provisioning, rebuilding, and API key management — not in the query path.
Recovery time objectives
Section titled “Recovery time objectives”| Scenario | Impact | Recovery |
|---|---|---|
| Running instance crashes | Queries unavailable for that instance | Instance is automatically restarted; typically recovers in under 2 minutes |
| Control plane API outage | Dashboard and provisioning unavailable; running instances unaffected | ReflexDB team resolves; check status.reflexdb.cloud |
| Instance data loss (e.g. infrastructure failure) | Queries unavailable | Re-provision from the dashboard — snapshot is rebuilt from your MySQL database |
What to do during an outage
Section titled “What to do during an outage”If your instance is down
Section titled “If your instance is down”- Check status.reflexdb.cloud for any active incidents
- Check the instance status in the dashboard — look for any error message or stuck build
- If the instance is in an error state, click Rebuild to trigger a fresh provision
- If rebuild fails or the instance is stuck, contact support@reflexdb.cloud with your database ID
If the dashboard or API is unavailable
Section titled “If the dashboard or API is unavailable”Running instances continue to serve queries. If you need to:
- Create a new API key while the dashboard is down: this requires the control plane — wait for recovery or use a management API key if you have one
- Provision a new instance: wait for the control plane to recover
- Query existing instances: no action needed — instances run independently
Contact support@reflexdb.cloud if the outage persists.
Re-provisioning an instance
Section titled “Re-provisioning an instance”If an instance needs to be rebuilt from scratch:
- Open the database in the dashboard
- Click Provision (if no instance exists) or Rebuild (if the instance is in an error state)
- The build takes 3–5 minutes; the instance will return to
runningautomatically - Existing API keys remain valid after a rebuild
Data residency
Section titled “Data residency”Synced data is held in memory on managed instances in your chosen region. It is:
- Never stored persistently by ReflexDB
- Never transmitted outside your selected region
- Rebuilt from your MySQL database on each provision or rebuild
If you terminate a ReflexDB instance, the in-memory snapshot is gone — but your MySQL data is unchanged.
Contact
Section titled “Contact”For incidents not covered here: support@reflexdb.cloud
Include your database ID (visible in the dashboard URL) and a description of the error or status you’re seeing.