Open dtest opened 9 years ago
That's not an unreasonable approach, but I get anxious about special snowflakes. I'd be happier to see us use groups more effectively. a low_cpu group and a high_cpu group for example.
Sensu groups might be the answer. My concern there is that it starts to grow fairly rapidly. For instance, I am thinking of an environment-based group (eg: prod-mysql1 might have different thresholds than qa-mysql1), which ends up being a lot of custom_json in the stack.
This would also break my mysql_graphite recipe because it depends on 'mysql' group subscription, unless we keep 'metric' plugins in general groups, but 'check' plugins in threshold-appropriate groups.
It would be great if we could override the sensu check thresholds setup by the custom_json on a per-client basis. I found this as an example: https://gist.github.com/piavlo/5774621
Is this the best way to do it?