L10
Recovery metrics and hot/warm/cold sites
Cram this first — memory hook
T = time to return; P = point to return to; Work Recovery Time = work catch-up; Maximum Tolerable Downtime = maximum business limit.
Exam trap
A hot site can still lose data if replication is inadequate. Recovery Point Objective is not how long the engineer takes to fix the server.
The slide facts
Recovery metrics include Recovery Time Objective , , and . The site comparison ranks hot, warm and cold by preparation, cost and restoration speed.
If you have time — extra
Recovery Time Objective () is the target time to restore a service. Recovery Point Objective () is how far back the recoverable data may be: the acceptable data-loss window. Work Recovery Time () is additional time to reconcile work and resume normal processing after technical restoration. Maximum Tolerable Downtime () is the business's limit; the course timeline places restoration and catch-up inside that limit. If a service must return within two hours and lose at most fifteen minutes of data, is two hours and is fifteen minutes. A hot site is highly prepared, a warm site partly prepared, and a cold site provides basic facilities requiring more setup. Think furnished house, partly furnished house, empty building. Specific recovery times depend on the design; the labels do not guarantee a number of minutes.
Walkthrough
Read the scenario for a downtime requirement, a data-loss requirement, and a budget. Assign / separately. Choose the site whose actual readiness supports the targets. Check replication, staffing and application dependencies too.