Select ALL measures from the lesson that can improve system availability.

### Why this is right - **Duplicate critical servers or storage components** is correct: redundancy can let another component take over when one fails. - **Distribute workloads across several resources** is correct: load balancing can prevent one resource from becoming overwhelmed. - **Monitor system metrics and alert administrators to problems** is correct: monitoring can reveal issues early enough for a response. - **Maintain backups and test disaster recovery** is correct: recovery plans and usable backups can restore data and services after disruption. ### The concept Availability controls reduce outages, limit overload, or help restore service. **Redundancy** provides duplicate components. **Load balancing** spreads work across multiple resources. Monitoring and alerts help operators detect problems, while backups and disaster recovery plans support restoration after failures or disasters. Controls should fit the system's needs and possible failure modes. Encryption protects data confidentiality in appropriate contexts, but it does not prevent traffic from consuming network capacity. ### The other options - **D. Encrypt network headers so routers cannot inspect them**: The lesson states that encryption cannot effectively protect against an availability attack such as a denial-of-service attack, and that encrypting routing information would limit routing or accessibility. ### Terms - **Redundancy**: Duplicating important components so another can serve if one fails. - **Load balancing**: Distributing work across resources to reduce overload on any one of them. - **Disaster recovery**: Planned actions for restoring data and services after a disruption. - **Encryption**: Transforming data into a protected form that requires a key to recover. ### Remember Availability is improved by preventing overload, tolerating component failure, detecting problems, and planning recovery.