You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
the K8s convention is to use camalCase and some of the fields in this API do that, right along others that use snake_case. This is a terrible user experience.
The text was updated successfully, but these errors were encountered:
I agree, it could be confusing. Proposed standard by kubernetes operator's guide - use camelCase, since operator-sdk uses JSON formatted tags.
But all VictoriaMetrics and prometheus configurations use yaml with snake_case. It makes impossible to copy-paste non-operator and operator based configs. It'll require an additional effort for that. For instance, scrape_interval from vmagent configuration docs could require to be converted into scrapeInterval.
My main pain point was AlertmanagerConfig. It's too verbose and it could be a nightmare to migrate from non-operator configuration to operator managed AlertmanagerConfig.
I think the good solution to go:
use snake_case for configuration options names exist at VictoriaMetrics yaml configurations. It should be straight forward. On other hand it makes CRDs incompatible with prometheus-operator format.
use camelCase for operator deployment params, like replicaCount or nodeSelector.
the K8s convention is to use
camalCase
and some of the fields in this API do that, right along others that usesnake_case
. This is a terrible user experience.The text was updated successfully, but these errors were encountered: