Everybody wants Azure

  • 1
    That's a good way to to conduct the disaster recovery exercise
  • 1
    Best resiliency testing:

    Kill arbitrary services at random, unplugging things if need be. Hungry interns chewing on cables should never kill your services, especially when they’re upstream.

    Best response: (persisted) queueing, exponential backoff, load warnings, etc.
  • 0
    @Root aka chaos testing
Add Comment