First thing I'd note is that we should use the OKD documentation. I guess that in most parts this is quite similar with the OCP version, but just to be on the safe side.
Comments are inline as well.
On 01.07.2019 17:43, Punga Dan wrote:
> Ran through similar issue/confusion. Apparently the openshift-metrics > ansible install which generates the openshift-infra project isn't't >
enough for the HPA to work. >
So you actually need to install the metrics-server for kubernetes which
> generates a new project called openshift-metrics-server inside your
Just to make sure that I have got you right: the installation notes on
both say that openshift-metrics/config.yml is the playbook to run. If I
understand you correctly then that is not true and I need to install
On both pages, OCP/OKD, for the Pod Autoscaling documentation you will find the reference to the metrics-server which needs to be installed. I haven't said anything about replacing the openshift-metrics project, but rather add it to the cluster. I guess more info about the different scopes of the 2 projects and how they interact and why we need them together could be shared by the guys from RedHat that know their use and integration.
Is that so because I used oc commands that are not meant for Openshift
3.11 or are the installation notes simply wrong?
I don't know about the "oc commands that are not meant for Openshift 3.11"...as long as you use the version 3.11 of the oc client, the commands should work fine. At keast they do for me...
About the installation notes/documentation I'd have to say that they can be improved. The reference about the metrics-server project and how to install it is only found on the development section page of Pod Autoscaling while there's no mention of it inside the cluster install/configure options.
Anyway, I would like to try what you recommended. How would I install
the metrics server for Kubernetes into m OpenShift cluster?
With the command documented on the Pod Autoscaling page. Both the OKD and OCP doc(which you've linked) states the same thing:
So invoke this playbook using the same inventory file you have used to install the cluster and be sure to supply the bold-ed variable.
users mailing list
users lists openshift redhat com