notes/Distributed_Systems/site_reliability_engineering.md

SRE (Site Reliability Enrineering)

Managing Machines

• Google Borg Ver15
  • similar to Apache Mesos
  • Kubernetes Bur16 (this is the successor)
• Clos network fabrik Clos53
• B4 - backbone network Jai13
• OpenFlow communication protocol

Storage

• Colossus
  • successor of a GFS (Google File System)Ghe03
  • comparable to Lustre or Hadoop Distributed File System (HDFS)

DB

• BigTable Cha06
• Spanner Cor12
• Blobstore

network

• Bandwidth Enforcer (BwE)Kum15
• Global Software Load Balacer (GSLB)
  • 3 levels: Geo DNS, user service, RPC

other

• Lock Service - Chubby Bur06
  • similar to Zookeeper
  • consistent DB
• Borgmon
  • similar to BB GUTS
• Stubby
  • similar to gRPC or BB BAS
  • uses protobufs, check also Apache Thrift

Dev Environment

Mor12b and Pot16