Kill Switches: The Safety Net Your Production Sync Workers Need
Why every production sync worker needs a kill switch, and how Automate Anything implements instant, granular control over every running sync operation.
In distributed systems, things can go wrong fast. A misconfigured sync worker can corrupt thousands of records before anyone notices. That's why we built kill switches into every sync worker on the platform.
**What Are Kill Switches?**
Kill switches are instant-off controls that stop a sync worker immediately, without waiting for current operations to complete. Unlike graceful shutdown, a kill switch is designed for emergencies.
**When to Use Kill Switches**
- Detecting unexpected data modifications at scale
- Third-party API returning corrupted data
- Runaway sync that's consuming excessive resources
- Security incident requiring immediate data flow stoppage
**Implementation Details**
Our kill switches operate at multiple levels:
1. **Worker-level**: Stop a specific sync worker instance
2. **Integration-level**: Pause all sync for a particular integration
3. **Account-level**: Emergency stop all sync operations
4. **Global-level**: Platform-wide pause (admin only)
Each level can be triggered via the dashboard, API, or automated rules.
**Recovery**
When a kill switch is activated, the system captures the exact state of all in-progress operations. When the switch is released, sync resumes from exactly where it left off, with no data loss or duplication.