As you prepare to deploy Tanzu Platform Self-Managed, you need to consider how you will use it so that you can make the appropriate decisions in how you install it. Specifically, the following aspects:
- What kind of environment do I have?
- How many apps and clusters/foundations will I manage?
Define your installation environment
You can install and run Tanzu Platform Self-Managed on a Kubernetes cluster in your own infrastructure in either an Internet-connected environment or an evironment that is completely disconnected from the Internet (air-gapped).
For an air-gapped installation, you need two machines:
- a jumpbox machine: an Internet-connected machine that you use to download the installer files
- a bootstrap machine: a disconnected machine that you use to perform the installation, and maintain the cluster, service, and registry
For an Internet-connected installation, you can use a single machine for both functions.
The Tanzu Platform installer supports deployment for both types of environment.
Topology of an air-gapped installation
Topology of an Internet-connected installation
Choose an installation flavor
The installation flavor defines which services and capabilities are available in the deployment. This allows you to tailor the installation for your needs. For this release, the following flavors are available:
- essentials This flavor installs Tanzu Platform, including features to support VMware Cloud Foundation inventory, Tanzu Platform for Cloud Foundry multi-foundation view, Tanzu Spring and Tanzu Salt.
- full This flavor installs Tanzu Platform with the essentials features, and additionally enables features to support Kubernetes operations, Spaces, and build services. This is the default mode for installation.
The installation flavor is defined in your installation configuration (config.yaml
) file.
Choose an installation profile
Your installation profile defines the resources required to deploy Tanzu Platform Self-Managed, which is correlated to how many resources you will be managing. It is possible to change profile later, if you need to expand your deployment. Select a profile that meets your current requirements and change profiles later as your requirements change.
For this release, you can choose from the following profiles:
- evaluation The evaluation profile has a minimal resource footprint, and can be used for demonstration and evaluation. It is not supported for production use. This profile allows you to create 2 projects with up to 10 application instances, and manage up to 20 Kubernetes clusters.
- foundation The foundation profile is a minimal production deployment, and intended for small environments or for customers who wish to gradually adopt Tanzu Platform and expand as needed. This profile allows you to create 4 projects with thousands of application instances, and manage up to 100 clusters.
- regular The regular profile corresponds to most customer environments. Most deployments use this profile because it provides a good balance between supported resources and hardware requirements. This profile allows you to create 6 projects with thousands of application instances, and manage up to 500 clusters.
- enterprise The enterprise profile is suitable for customers who require support for thousands of applications and resources. Select this profile initially only if you expect to support more resources than the regular profile. This profile allows you to create 6 projects with thousands of application instances, and manage up to 1000 clusters.
The following table shows the specific resource requirements and load limits of the available profiles.
Installation Profile | Usage Limits | Observability Load | Resource Requirements |
evaluation | Ingest: 20 RPS Query: 2 QPS Concurrent assessments: 2 Application instances: 10 Resource dimension: 2 projects, 2 spaces, 20 clusters |
Metrics PPS : 1k Log messages per second: 500 |
Control plane Nodes: 3 vCPUs per node: 4 Memory per node: 16GB Storage per node: 20GB Workers Nodes: 3 vCPUs per node: 8 Memory per node: 32 GB Storage per node: 40 GB |
foundation | Ingest: 100 RPS Query: 20 QPS Concurrent assessments: 50 Application instances: 5000 Resource dimension: 4 projects, 10 spaces, 100 clusters |
Metrics PPS : 10k Log messages per second: 5k |
Control plane Nodes: 3 vCPUs per node: 4 Memory per node: 16GB Storage per node: 20GB Workers Nodes: 9 vCPUs per node: 8 Memory per node: 32 GB Storage per node: 40 GB |
regular | Ingest: 200 RPS Query: 50 QPS Concurrent assessments: 100 Application instances: 15000 Resource dimension: 6 projects, 20 spaces, 500 clusters |
Metrics PPS : 20k Log messages per second: 10k |
Control plane Nodes: 3 vCPUs per node: 4 Memory per node: 16GB Storage per node: 20GB Workers Nodes: 10 vCPUs per node: 12 Memory per node: 48 GB Storage per node: 40 GB |
enterprise | Ingest: 500 RPS Query: 100 QPS Concurrent assessments: 200 Application instances: 30000 Resource dimension: 6 projects, 30 spaces, 1000 clusters |
Metrics PPS : 40k Log messages per second: 20k |
Control plane Nodes: 3 vCPUs per node: 4 Memory per node: 16GB Storage per node: 20GB Workers Nodes: 11 vCPUs per node: 16 Memory per node: 64 GB Storage per node: 40 GB |
For more help with planning your Tanzu Platform Self-Managed deployment, see the Tanzu Platform Self-Managed on vSphere - Dual-Site Reference Design.
Content feedback and comments