Dell ObjectScale Application 1.3.x Installation Guide for Red Hat OpenShift

CSI Secondary scheduler and its namespace

Also, ObjectScale requires the use of a CSI secondary-scheduler operator. During the installation process that follows you deploy the secondary-scheduler operator into a separate secondary-scheduler operator namespace.

Bare-Metal CSI Driver and its namespace

ObjectScale on OpenShift requires the use of the Bare-Metal CSI Driver. During the installation process that follows you deploy the Bare-Metal CSI Driver into a separate CSI Bare-Metal namespace.

ObjectScale and object store namespaces

ObjectScale and its object store can either both be deployed in the same or in separate nondefault OpenShift namespaces. During the installation process that follows you deploy ObjectScale into its own namespace.

• If you are deploying ObjectScale and the object stores in different namespaces within the Kubernetes cluster, here is one way of naming and organizing the cluster namespaces:

  • Create a project objectscale-system for the ObjectScale namespace.
  • Create one or more object-store-x projects as other namespaces where you can deploy all the object store resources.
• As the Admin user, assign RBAC roles to the users so that they can manage the specified namespace and the object stores.
  NOTE:When creating end users, only assign these users namespace-level role bindings. ObjectScale on OpenShift only supports Kubernetes-level namespace bindings and does not support cluster-level role bindings.

OpenShift users and ObjectScale

An OpenShift user with the cluster-admin role, who has access to all the namespaces in the OpenShift cluster, can log in to the OpenShift UI and create projects, namespaces for ObjectScale installation, and the end users for ObjectScale. These end users cannot be assigned the cluster-admin role. Rather, they need the admin role for the namespaces where ObjectScale and object stores are deployed.

The cluster-admin user can also optionally create a separate, end-user namespace, where end users can create and manage end-user object stores.

Before deploying ObjectScale on the OpenShift cluster, you can complete the following actions within the OpenShift UI following the Red Hat OpenShift documentation:

• Log in to the OpenShift UI as kubeadmin.
• As the kubeadmin user, create users and groups by configuring one of the identity providers that is supported on the OpenShift platform.
• Then, assign the cluster-admin role to one of the users so that it can manage the cluster as admin while logged in as a user.
• As the Admin user, create the necessary projects (namespaces) in the Kubernetes cluster to manage the ObjectScale deployment. You have the option of deploying the ObjectScale instance and the object stores within the same or different namespaces.
• As the Admin user, assign RBAC roles to ObjectScale and object store users in OpenShift so that they can manage the specified namespace and object stores.
  NOTE:When creating end users, only assign these users namespace-level role bindings. ObjectScale on OpenShift only supports Kubernetes-level namespace bindings and does not support cluster-level role bindings.