CompTIA Network+ Preparation · lesson 24 of 30 · 25 min
High Availability Concepts
High Availability (HA) — designing systems to minimize downtime. Measured as a percentage of uptime. “Five nines” (99.999%) = ~5 minutes downtime per year.
SPOF (Single Point of Failure) — any component whose failure brings down the entire system. HA design eliminates SPOFs through redundancy.
Redundancy — duplicate components so failure of one does not cause an outage. Applies to power supplies, NICs, switches, links, ISPs, and data centers.
Load balancing — distributes traffic across multiple servers. Improves performance and provides redundancy. If one server fails, others handle the load. Methods: round-robin, least connections, IP hash.
Clustering — multiple servers working together as one. Active-active: all nodes handle traffic. Active-passive: standby node takes over on failure.
NIC teaming (bonding) — combining multiple NICs for redundancy and/or increased bandwidth. LACP (802.3ad) is the standard protocol for link aggregation.
Disaster Recovery Metrics
RTO (Recovery Time Objective) — maximum acceptable time to restore a system after a failure. “How long can we be down?” Lower RTO = more expensive solution.
RPO (Recovery Point Objective) — maximum acceptable data loss measured in time. “How much data can we afford to lose?” RPO of 1 hour means backups every hour.
MTTR (Mean Time to Repair) — average time to restore a failed component. Lower is better.
MTBF (Mean Time Between Failures) — average time between failures. Higher is better — indicates reliability.
Backup and Recovery Sites
Hot site — fully operational duplicate data center with live data replication. Can take over in minutes. Most expensive.
Warm site — has hardware and connectivity but data must be restored from backup. Recovery takes hours to days. Moderate cost.
Cold site — empty facility with power and connectivity only. Hardware must be installed and data restored. Recovery takes days to weeks. Cheapest.
Backup types:
Full backup — copies all data. Slowest to create, fastest to restore.
Incremental backup — copies only data changed since the last backup (full or incremental). Fastest to create, slowest to restore (need full + all incrementals).
Differential backup — copies all data changed since the last full backup. Moderate create time, faster restore than incremental (need full + latest differential only).
Exam Tip
RTO = downtime tolerance. RPO = data loss tolerance. Hot site = minutes to recover. Cold site = days. Incremental = fastest backup, slowest restore. Differential = moderate backup, faster restore than incremental.
Key Terms
Key Terms
Knowledge Check — 10 Questions
Network+ — Availability and DR
RTO defines:
- How much data can be lost
- Maximum acceptable downtime after a failure
- Average time between failures
- Time to complete a backup
- Answer: Maximum acceptable downtime after a failure
RTO = Recovery Time Objective — how long the business can tolerate being down. Lower RTO requires more expensive HA solutions. RPO of 4 hours means:
- Systems must be restored within 4 hours
- Backups must occur at least every 4 hours
- The system has 4 hours of battery backup
- MTTR must be under 4 hours
- Answer: Backups must occur at least every 4 hours
RPO = Recovery Point Objective — maximum data loss tolerance. RPO of 4 hours means you can lose at most 4 hours of data, so backups every 4 hours. A hot site can restore operations in:
- Days to weeks
- Hours to days
- Minutes
- Seconds only
- Answer: Minutes
Hot site = fully operational duplicate with live data replication. Failover can happen in minutes. Most expensive DR option.
Which backup type is fastest to create but slowest to restore?
- Full
- Differential
- Incremental
- Mirror
Answer: Incremental
Incremental backups only copy changes since the last backup — fast to create. Restore requires the full backup plus every incremental — slow.
LACP (802.3ad) is used for:
- Wireless authentication
- Link aggregation / NIC teaming
- VLAN tagging
- Spanning tree
- Answer: Link aggregation / NIC teaming
LACP (Link Aggregation Control Protocol) bundles multiple physical links into one logical link for redundancy and bandwidth. MTBF measures:
- Time to restore after failure
- Average time between failures — reliability indicator
- Time to complete a backup
- Maximum downtime allowed
- Answer: Average time between failures — reliability indicator
MTBF = Mean Time Between Failures. Higher MTBF = more reliable component. Used to compare hardware reliability. A cold site provides:
- Live data replication
- Pre-installed hardware ready to use
- Empty facility with power and connectivity only
- Automatic failover
- Answer: Empty facility with power and connectivity only
Cold site = empty shell. Hardware must be installed and data restored from backup. Recovery takes days to weeks. Cheapest option. Active-active clustering means:
- One node is always on standby
- All nodes actively handle traffic simultaneously
- Nodes alternate daily
- Only one node runs at a time
- Answer: All nodes actively handle traffic simultaneously
Active-active: all cluster nodes handle traffic — better performance and redundancy. Active-passive: standby node only activates on failure. Differential backup copies:
- All data
- Only changes since the last full backup
- Only changes since the last backup of any type
- Only deleted files
- Answer: Only changes since the last full backup
Differential: copies everything changed since the last full backup. Restore = full + latest differential only (faster than incremental restore).
Five nines availability (99.999%) allows approximately _____ minutes of downtime per year.
Answer: 5
99.999% uptime = 0.001% downtime. 0.00001 × 525,600 minutes/year ≈ 5.26 minutes. “Five nines” is the gold standard for HA.
Previous lesson: Wireless security threats: evil twin, deauthentication, WPS attacks, rogue APs
Back to CompTIA Network+ Preparation
Next lesson: VPN types: site-to-site, remote access, SSL/TLS, IPsec, split tunneling