Distributed Logging and Monitoring

• What we get:

  • Unified logging for local area clusters
  • No drop semantics
  • Multiple subscribers with little overhead
    • Fault-tolerant log writers
    • Process real-time cross-server clickstreams
    • Real-time per server and cluster-wide hit analysis
    • Real-time individualized mass vhost logging
    • ALL AT THE SAME TIME!

Notes: