Configure Openshift Monitoring Best Practices | Known issues

Products

DX Unified Infrastructure Management (Nimsoft / UIM)

Issue/Introduction

Setting up the new version of DX UIM probe Openshift 2.01 / 3.00 and not able to connect | "Profile in failure state"
The probe new version is restructured in comparison to the previous version. What is required now to deploy the OpenShift Monitoring Configuration?
The documentation for Deploying DX UIM Images for OpenShift Monitoring doesn't mention the need to provide a pull secret. Is this needed for deploying the images like in earlier versions?
The documentation doesn't mention the need to customize the config map with local cluster details. Is this needed for connection to occur?
After successful deployments (pods are running, Connection to cluster was working, Metrics were being collected) but all metric values are 0.00
What endpoints/ports need to be enabled between the probe and the cluster in comparison to previous versions?
What is the difference between Openshift version 2.01 and 3.00? Should I upgrade to 3.00?

Environment

DX UIM 23.4.*
Openshift probe 2.01 / 3.00

Cause

Documentation Deficiency, Best Practices, Known issues

Resolution

Pull Secret

The OpenShift probe images is pulled from Docker Hub and not from the Broadcom container registry. Therefore, a pull secret is not be required. The OpenShift image comes from the Docker Hub repository for the Universal Monitoring Agent. Since these are public Docker Hub images, anonymous pulls are allowed and OpenShift is able to pull the image without credentials.

Create a profile, Endpoint connectivity Requirements

In comparison to older version of the probe new end points are required to be reachable from the OpenShift probe system. This is especially important if there is a load balancer in the middle.
Specifically the following is needed to be open between the probe and the cluster:

Prometheus Exposed Port (Default port 31090):
Specify the port number to access the metrics from the Prometheus importer. Run the following command to get the port.

oc get services -n prometheus

Collector Exposed Port (Default port 30779)
Specify the port number to access the APIs of the collector service, which is deployed on the master node.

Port: (Default port 30778)
Specify the clusterinfo port number used to access the OpenShift config that monitors the API URL.
Run the following command to get the port:
oc get services -n dxuim

Collecting metrics with value 0.00: Update the image pulled by default

After successful deployments (pods are running, connection to cluster was working, Metrics were being collected) but all metric values may still be collecting at 0.00 vale.
This occurs because the default image or the one pulled via generic tags does not contain the necessary fixes for metric processing logic introduced in the latest probe iterations.

To resolve this issue:

Edit the Helm values file:

uim/values.yaml
Update the image version:

imageName: caapm/universalmonitoragent:26.1.1.35
Apply the upgrade:

helm upgrade uim ./uim/ -n dxuim
Verify that pods restart and metrics begin reporting non-zero values.

NOTE: an additional issue was addressed that caused metrics to show as 0.00, the fix can be downloaded in the KB --> Some Openshift metrics collect metrics value as 0.0

Openshift version 3.00:

Openshift probe v3.0 was published to provide support for Java21. It is recommended to upgrade to version 3.00 if upgrading to DX UIM 23.4.5 (CU5) or later.

This version of probe doesn't any major changes in comparison to 2.01 see: Openshift Release Notes

Additional Information

Additional Notes as of As of March 2026:

The official documentation located here OpenShift Monitoring Configuration
A new commons_container package will be published that include new image tag out of the box.