There’s an interesting thing about incidents, once they’re resolved they no longer need attention, especially not attention on how they might be resolved better in future.
But there’s something of the special snowflake about incidents – they’re all just a little bit different or unique and need their own special attention. [I’m not talking about the “noise” incidents like flapping ports here of course].
With every incident solved, you could say that an opportunity for improvement and human teaching is lost.
What if we were able to accurately record all of the context leading up to an incident and play it back, allowing many people to try to resolve it and noting the most efficient resolution with some sort of scoreboard (eg fastest resolution times, most accurate resolution of root cause, etc), thus gamifying?
We could crowd-resolve it. We could teach the most efficient resolution. We could test new hires to see how efficient they are.
Could it even make our machine intelligence even more intelligent?Read the Passionate About OSS blog for more.