English
Menu

System Architecture and Reliability

For a system, the RPO (recovery point objective) is set at 4 hours and the RTO (recovery time objective) at 2 hours. A failure can occur at any time, and the time taken to make a backup is negligible. Which data protection scheme meets these targets?

  1. ABackups are taken every 1 hour, and recovery takes 3 hours when a failure occurs
  2. BBackups are taken every 3 hours, and recovery takes 1.5 hours when a failure occurs
  3. CBackups are taken every 6 hours, and recovery takes 1 hour when a failure occurs
Show answer and explanation

Answer: B. Backups are taken every 3 hours, and recovery takes 1.5 hours when a failure occurs

If a failure occurs just before the next backup, data for the whole backup interval is lost, so meeting the RPO requires a backup interval no longer than the RPO. Also, the recovery time must be no longer than the RTO. With backups every 3 hours and recovery in 1.5 hours, at most 3 hours of data is lost (4 hours or less) and recovery takes 1.5 hours (2 hours or less), so both are met. The 1-hour scheme fails the RTO, and the 6-hour scheme fails the RPO.

  • A. It meets the RPO, but recovery takes 3 hours, exceeding the 2-hour RTO.
  • B. Correct. Data loss is at most 3 hours and recovery takes 1.5 hours, both within the targets.
  • C. Recovery is fast, but up to 6 hours of data is lost, exceeding the 4-hour RPO.
Practice this topic Back to topic

Related calculators

RAID Capacity CalculatorCalculate usable capacity, capacity efficiency, how many disk failures can be tolerated and the chance of data loss for RAID 0, 1, 5, 6 and 10 from the number and size of disksAvailability CalculatorCompute availability from MTBF and MTTR, and the availability and annual downtime of systems built from series, parallel and k-out-of-n units

Books on System Architecture and Reliability Ad

As an Amazon Associate I earn from qualifying purchases.

More System Architecture and Reliability questions

  1. Database updates are replicated synchronously between two data centers 1,000 km apart. For each update, the system waits for one round trip: the transfer to the other site and the notification that the write is complete. Taking the speed of light in optical fiber as about 200,000 km/s and the fiber route length as equal to the straight-line distance, which is closest to the increase in the response time of one update caused by propagation alone?
  2. In an IaaS cloud service, which does the user generally manage?
  3. Among the ways to increase a system's processing capacity, which corresponds to scale-out?
  4. Which is generally a desirable design for a system that uses autoscaling to add and remove servers automatically according to the load?
  5. Which describes live migration in a virtualized environment?