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
I would like to suggest the addition of a Keptn metrics provider to the list of already implemented providers.
I believe that, given the metric management capabilities of Keptn, would make this a valuable addition to Flagger and can benefit both projects. Being one of the Keptn maintainers, I would be happy to work on the implementation and documentation for this feature if this is something you would like to have added to Flagger :)
Proposed solution
I implemented a quick proof of concept in my fork of the flagger repo here.
The main difference to other providers in Flagger is that this one is using a Kubernetes client to interact with Keptn resources (KeptnMetrics and Analyses). Therefore, in addition to the code in keptn.go, the ClusterRole needs to be adapted to be able to access the required Keptn resources.
The provider supports the use of the following resources:
KeptnMetric: This resource represents the current value of a certain metric. The metric is retrieved by Keptn from one of the following data sources: prometheus, thanos, cortex, dynatrace, datadog.
Analysis: This Keptn resource can be used to calculate an overall score based multiple metric, and define a threshold for the overall score to be considered as successful. More information about this concept can be found in the Keptn Docs
The text was updated successfully, but these errors were encountered:
Hi @bacherfl I'm happy to have this integration in Flagger. I see that the API version is at v1beta1, would you be able to help maintain the integration in Flagger going forward?
Describe the feature
I would like to suggest the addition of a Keptn metrics provider to the list of already implemented providers.
I believe that, given the metric management capabilities of Keptn, would make this a valuable addition to Flagger and can benefit both projects. Being one of the Keptn maintainers, I would be happy to work on the implementation and documentation for this feature if this is something you would like to have added to Flagger :)
Proposed solution
I implemented a quick proof of concept in my fork of the flagger repo here.
The main difference to other providers in Flagger is that this one is using a Kubernetes client to interact with Keptn resources (
KeptnMetrics
andAnalyses
). Therefore, in addition to the code inkeptn.go
, theClusterRole
needs to be adapted to be able to access the required Keptn resources.The provider supports the use of the following resources:
KeptnMetric
: This resource represents the current value of a certain metric. The metric is retrieved by Keptn from one of the following data sources:prometheus
,thanos
,cortex
,dynatrace
,datadog
.Analysis
: This Keptn resource can be used to calculate an overall score based multiple metric, and define a threshold for the overall score to be considered as successful. More information about this concept can be found in the Keptn DocsThe text was updated successfully, but these errors were encountered: