I have a question related to openshift pod tolerations. According to the document, tolerations are defined in pod. Spec not in deployment/deploymentconfigs Here is the scenario, 3 master nodes with following taints, no other nodes oc adm taint node master01 want=food:NoSchedule oc adm taint node master02 want=drink:NoSchedule oc adm taint node master03 want=money:NoSchedule After deploying a simple hello-world app, not a surprise that pod stucks in pending state because pod doesn't have any tolerations Name READY STATUS RESTARTS AGE hello-world-1-build 0/1 Pending 0 5s Now add a toleration that tolerates all taints to hello-world-1-build tolerations: - operator: Exists Now build is running but the application pod still in pending state, because I only added toleration to build pods. Adding the same tolerations to hello-world pod will change the state to running but if I want to scale to 10 pods, then I would have to manually add tolerations to all 10 pods. Is there a better way, without removing the node taints?

I'm working with customers who require the configuration of dedicated CPUs in OpenShift. While I’m familiar with options like CPU isolation and reserved CPU cores, I’m looking for a way to pin specific CPU cores to a pod. The goal is to ensure that these CPU cores are fully dedicated to the pod and not accessible to other pods or processes running on the operating system. I'm also familiar with taskset/cset and other linux commands, but need this to be addressed on the OpenShift/K8s level. Is there a way to achieve this in OpenShift, ensuring that the CPU core remains exclusive to the assigned pod without any interference from other processes? UPDATE: I’m looking for exclusive cpu usage. Without the option for system services of the node or other non pinned pods catching the entire vcpus of the machine. For example, if I’m asking for 4 CPU’s. Let’s assume I get 2,3,4,5 out of a 24 core node. I would like to avoid pods or aervices without pinning to get all cpus (0-23) which will also get 2,3,4,5.

I'm looking at deploying Ansible Tower on OpenShift but don't want to deal with deploying my own Postgresql instance if I don't have to (not that it's a big deal). If I deploy [Ansible Tower](https://docs.ansible.com/ansible-tower/latest/html/administration/openshift_configuration.html#ag-openshift-configuration) on OpenShift, the diagram makes it look like I have to deploy my own instance of Postgresql instead of it using its normal embedded instance. Is that accurate or will it spin up its own Postgresql pod?

I have pods running in OpenShift that I would like to view the logs for. In particular, I want to see the logs when the pod is terminated. However, when the pod is terminated the logs are deleted as far as I can tell. I've tried tailing 'oc logs ' but I lose the output once the pod is gone. What is a clean way to copy these logs in real time or print them to stdout so that they will persist after the pod is killed?

I am downloading some RPM packages from the Red Hat Customer site and came across some rhaos packages. I'm not sure what they are exactly. For example: podman-4.4.1-3.rhaos4.13.el8.x86_64.rpm I couldn't find much information about them, except that they are related to OpenShift. Can these packages be safely installed on RHEL?