Provisioning compute resources
Multitenant compute resources, such as vCPUs, RAM, storage, floating IPs, load balances and Kubernetes clusters, can be provisioned to end users in the self-service panel. Additionally, you can install the billing metering service to collect usage data of compute resources in different projects.
Limitations
- Compute storage space is not fully thin provisioned. After user data removal, unused storage space is not reclaimed and is reported as actual used space, which is charged according to your licensing model. For more details, refer to Logical space chart.
Prerequisites
- A clear understanding of the compute cluster, which is explained in About the compute cluster.
- Your hardware meets the requirements listed in Compute cluster requirements.
- Your infrastructure networks are set up, as described in Setting up networks for the compute cluster.
- The storage cluster is created by following the instructions in Deploying the storage cluster.
- Domains, projects, and users are created, as described in Configuring multitenancy.
- You have access to the self-service panel configured, as outlined in Providing access to the self-service portal.
Provisioning overview
- Create the compute cluster. After the cluster is created, you can manage compute objects in the admin panel (refer to Managing the compute cluster).
- Set the same CPU model to all of the compute nodes.
- Set a DNS name for the compute API.
- Secure the OpenStack API traffic with an SSL certificate.
- If you want to provision Kubernetes as a Service, install the Kubernetes service.
- If you want to provision Load Balancer as a Service, install the load balancer service.
- If you want to provision billing metering, install the billing metering service.
- If you want to provision Backup and Restore as a Service, enable and configure the backup service.
- To manage compute resources in the self-service panel, log in there at http://<admin_panel_IP_address>:8800 as a self-service user, and then proceed to create virtual objects, as described in Managing compute resources in the Self-Service Guide.