FAULT-TOLERANT COMPUTING Failure(失效/故障)：Vhen a component is not living up to its specifications,a failure occurs Error(错误)：The part of a component's state that can lead to a failure Fault(缺陷/故障)：The cause of an error.Types: Transient(偶发)：occur once,then disappear -Intermittent(间歇：occur,.then vanish,then reappear -Permanent(持久)：continues to exist 指标 -1 可靠性：MTBF=MTTF+MTTR,年失效率AFR -可用性=MTTF/(MTTF+MTTR)=工作时间/（工作时间+维修时间） 9°s Availability Downtime/Year Examples MTBF 1 90.0% 36 days 12 hours Personal clients 2 99.0% 87 hours 36 minutes Entry-level businesses MTTF MTTD MTTR MTTF 3 99.9% 8 hours 46 minutes ISPs,mainstream businesses Correct behavior Diagnose Repair Correct behavior Time 4 99.99% 52 minutes 33 seconds Data centers 5 99.999% 5 minutes 15 seconds Carrier-grade Telco,medical, First Begin End Second banking Failure Repair Repair Failure 99.9999% 31.5 seconds Military defense systemFAULT-TOLERANT COMPUTING • Failure(失效/故障): When a component is not living up to its specifications, a failure occurs • Error(错误): The part of a component's state that can lead to a failure • Fault(缺陷/故障): The cause of an error. Types: – Transient(偶发): occur once, then disappear – Intermittent(间歇): occur, then vanish, then reappear – Permanent(持久): continues to exist • 指标 – 可靠性：MTBF = MTTF+MTTR，年失效率AFR – 可用性=MTTF/(MTTF+MTTR)=工作时间/（工作时间+维修时间）