Objective.
Kill a cache instance and measure how long until it answers again.
Scenario.
The cache instance was killed without warning.
Expected behaviour.
- Restart automatically.
- Answer again within seconds.
Measured recovery.
- Recovery time RTO
- 1.2 s
- How long recovery took.
- Application health
- Pass
Customer impact.
- Application availability
- Answering again after 1.2 s
What happened.
- Cache instance killed.
- Answering again.
What we learned.
- The cache is back in just over a second.
What we changed.
- No change needed.
Scope of this drill.
One cache instance.