What is an Emergency Pause?
An emergency pause is a safety mechanism that allows designated parties to halt some or all protocol operations when a security threat is detected. This provides time to assess, respond, and fix issues before damage spreads.
Why Emergency Pause Exists
DeFi protocols can't patch bugs instantly like web applications. An emergency pause:
- Stops ongoing exploits
- Prevents further losses
- Provides response time
- Protects user funds
What Gets Paused
Different protocols pause different functions:
- Full Pause: All operations stop
- Partial Pause: Only affected functions paused
- Per-Market Pause: Individual asset pauses
- Directional Pause: Block deposits but allow withdrawals
Who Can Pause
Guardian: Single trusted party (fastest response) Multisig: Group consensus required Governance: Token holder vote (slowest but most decentralized) Automated: Code-triggered based on conditionsEmergency Pause Patterns
Pausable Contract: OpenZeppelin's standard implementation with whenNotPaused modifier. Per-Function Pause: Granular control over operations Time-Limited Pause: Auto-unpause after duration Escalating Pause: Guardian can pause, governance must unpauseReal-World Pause Events
- Compound: Paused during oracle manipulation incident
- Aave: Market-specific pauses during volatility
- Curve: Pool pauses during LP token pricing issues
- Bridges: Paused after hack detection
Pause Duration Considerations
Short pauses: Contain immediate threat but limited analysis time
Long pauses: Thorough investigation but users locked out
Best practice: Clear unpause criteria and timeline communication
User Impact During Pause
Pauses affect users differently:
- Can't deposit or withdraw
- Positions may be at risk
- Liquidations may be blocked (good or bad)
- Interest may or may not accrue
Pause as Centralization
Emergency pause is inherent centralization:
- Creates trusted party dependency
- Could be abused for censorship
- Must be balanced against security needs
- Progressive decentralization of pause rights
Best Practices
- Clear communication during pauses
- Transparent trigger criteria
- Defined response procedures
- Regular pause mechanism testing
- Post-incident analysis and reporting