Closed lmilbaum closed 4 years ago
A short related presentation: https://docs.google.com/presentation/d/1yGiGz9NNBYaIlnDj9uhFdk90vnUzELgGKopk7jh7uOc/edit#slide=id.g4baccfa9d2_1_24
Are we taking care of spawning the environment or starting from an existing environment?
I would vote for taking care of the spawning as it makes it easier to test the project in the future + best customer experience :D
Delivery Pipeline is a multi sub-services service. Each sub-service may through alerts when hitting a failure or close to hitting one.
The alerts should be pushed to a one and only one queue. This is the first sub-service.
The queue should be able to:
Consume the alerts and transform them to incidents
Analyze the alerts - removing false/positives and aggregation of similar alerts
Tracking incidents are resolved according to SLA
Remediation by automation