OpenShift HAProxy ingress monitoring

[joseandrescalvo]

Hi.

We have an OpenShift cluster successfully monitored with Dynatrace Operator. We are trying to monitor the OpenShift HAProxy deployed using Ingress Operator (https://docs.openshift.com/container-platform/4.12/networking/ingress-operator.html). This process does not support full-stack monitoring so the only option is to get metrics exposed with Prometheus. We have tried the following approaches:

• The extension HAProxy (Prometheus) seems to not work when installing in OpenShift.
• The HAProxy pod seems to be managed by the Ingress Operator, so annotations can not be added in order to scrape metrics with the automatic mechanism in Dynatrace to get Prometheus metrics: https://www.dynatrace.com/support/help/shortlink/monitor-prometheus-metrics.

Any suggestion?

[joseandrescalvo]

Additional info about this issue:

• The HAProxy Prometheus extension does not work when deployed in OpenShift. Response from Dynatrace support:

This extension as well as any other extension 2.0 cannot be deployed in OpenShift or Kubernetes, because it is not supported yet. In other words you cannot deploy any extension to containerized OneAgent or ActiveGate. For now it is only possible to deploy HAProxy (Prometheus) extension to non-containerized OneAgent or ActiveGate (outside of Kubernetes/OpenShift). 

• The extension can remote monitoring HAProxy (Prometheus) but data is not associated with a Host and so consumes DDUs in all cases (metrics not deducted from hosts' free tier).

HTH