智能运维中的异常检测和根源分析.pdf

• • • AI + • • • metrics • Block Time (ms), Average Block Time (ms) • Calls, Number of Calls • Calls/min, Calls per Minute • CPU Used (ms), JVM CPU Burnt (ms/min) • Errors/min, Errors per Minute • Response Time (ms), Average Response Time (ms), Avg. Time per Call • Slow Transactions, Number of Slow Calls • Stalled Transactions, Stall Count • Wait Time (ms), Average Wait time (ms) • Backends • End User Monitoring UE • Mobile • Service End Points • Overall Application Performance • Business Transaction Performance • Application Infrastructure Performance • Errors • observation • min max • sum • avg • count • % • percentile • • • AI + • • • • value X • value Y • 1μ 2σ • μ ± 2σ • value 1: 
 seasonality