谷歌运维解密翻译作者讲解SRE

admin

: ~; K8 D6 i% u9 J• 生产线管理员
• Ensure user-visible uptime and service quality
• Authority over production environment.
• 跟网站一起成长
' a6 \3 J% N3 A- R) a+ N1 o" t• Steep learning curve, mostly due to complexity
. `: T8 [8 n% ^6 A• Continuous retraining, sites always being improved
6 t' E; G* `2 q$ j, F& P6 m• 基础架构设施
• Specializations for shared infrastructure
• Ensure those components have good reliability

. `  [1 T" B8 X  Z
" a: ?: u1 b. H% ^5 v2 n& ?3 _it just works
- q" h4 k! U# f  O, H3 M3 i• Service Level Objective (SLO)
• Monitoring/Deployment
1 q# Y$ B" {+ Z# j- a5 v5 F: G• Capacity Planning
, i( Y6 s% _( m  M& c# x( b• 以一敌百
• Team manages monitoring and develops automation
2 E* K: w8 c# c; h% k( I$ V) b• Implies use of scripting and data analysis tools
2 G$ [8 X3 d' \$ w0 e& \• Most failures need automated recoveries in place
3 O. A' R6 q6 E( E+ c. q( c• 救火队员和纵火犯合体
5 j, z! A3 T8 ]9 [) }! z) r• Elevated risk during convenient working hours
• Learn of age mortality risk during preceding workday
• Infant mortality ideally also avoids meals

. B5 p& E  v5 o, z5 |
9 i. }4 j0 }  Q码农
5 |4 z2 V! M: M. h" ]' U• Not administration
• 报警系统重度（中毒）用户
- H3 E! V* B2 ?5 m% M, o2 j! }• Holes may cause outage before notification occurs
• Routinely use multiple layers, levels and viewpoints
0 Y( t; ^4 Z& w; l3 X" z7 ^( A• Design the manual and automatic escalation paths
• 对未来负责
• Responsible for enabling growth and scaling
• Plan for requirements, identify inefficiencies
• File bugs and, where appropriate, fix them too
, f" ~% u4 u- m1 |  C1 K
8 D4 \# ~; s- k3 G, b
" j. |, P; d/ \ QCon上海2015-来自-Google-DevOps-经验的落地实践-SRE.pdf (469.17 KB, 下载次数: 0)

2020-11-20 12:11 上传

点击文件名下载附件
0 r, [  W# F2 {