Alright... this seems somewhat aligned with the Observability 2.0 work, but a conversation the other day resparked the interest in the ideal (see #33 for more context). When I was working at an un-named company, one thing we talked about was being able to write an "SLO language" where we could define our SLO's using a common language. If I think about when this conversation was happening, there weren't great foundations for building such a system, but I think that has changed today. These days, writing a custom scripting language using a typescript-esq DSL is fairly straightforward thanks to solutions like Deno. The Aurae project did this as part of their specification language and I think this might just be the key to actually implementing a solution...
This issue would address some of the problems in #10, and could even obviate it to some extent by replacing it with a more useful solution.
Ideally, I want to be able to describe a given service level objective and have the resulting metric queries, alerts, and graphs all be generated for me... the main idea here is to make writing this more approachable, without application developers needing to know the intricacies of your organizations infrastructure.
Alright... this seems somewhat aligned with the Observability 2.0 work, but a conversation the other day resparked the interest in the ideal (see #33 for more context). When I was working at an un-named company, one thing we talked about was being able to write an "SLO language" where we could define our SLO's using a common language. If I think about when this conversation was happening, there weren't great foundations for building such a system, but I think that has changed today. These days, writing a custom scripting language using a typescript-esq DSL is fairly straightforward thanks to solutions like Deno. The Aurae project did this as part of their specification language and I think this might just be the key to actually implementing a solution...
https://github.com/aurae-runtime/aurae/tree/main/auraescript
This issue would address some of the problems in #10, and could even obviate it to some extent by replacing it with a more useful solution.
Ideally, I want to be able to describe a given service level objective and have the resulting metric queries, alerts, and graphs all be generated for me... the main idea here is to make writing this more approachable, without application developers needing to know the intricacies of your organizations infrastructure.