The Paramify Platform Installer (PPI)

Overview

The Paramify Platform Installer (PPI) is an administrative console where you can access and manage your Paramify instance. It is also where you will go to access upgrades to new Paramify releases, as well as any updates to the installer itself. The installer consists of an Application tab where you manage your application, version history, and configuration settings.

TLDR;​

1. Run the install script and wait.

curl https://kots.io/install | bash
kubectl kots install paramify

2. Choose an admin console password.
3. Upload your license file.
4. Enter custom configuration details.
5. Deploy the application.

System Requirements​

Here are the system requirements for installing Paramify on an existing Kubernetes cluster:

• Paramify currently only supports deploying onto amd64/x86_64 based architectures. Deploying onto arm based architectures is not yet supported.
• A Kubernetes 1.19.x, 1.20.x, 1.21.x, 1.22.x, 1.23.x, 1.24.x, 1.25.x, 1.26.x, or 1.27.x compliant cluster.
  • Paramify has tested and verified compatibility with the following managed Kubernetes providers:
    • Amazon EKS (1.23, 1.24, 1.25, 1.26, 1.27)
    • Google GKE (1.25, 1.26, 1.27)
    • K3S (1.24, 1.25, 1.26)
  • Other distributions, such as Azure AKS, may or may not be compatible with Paramify.
• 3+ node cluster for production environments to provide redundancy and high availability (HA).
  • Ensure all nodes are in the same cloud provider region or physical data center network. Nodes behind different WAN links in the same cluster are not supported.
• Node resources:
  • CPU:
    • 4 cores. The cluster must contain at least 2 cores, and should contain at least 4 cores.
  • Memory:
    • 1Gi free memory. The cluster must have at least 512Mi free memory, and should have at least 1Gi free memory.
  • Storage:
    • 20Gi free disk space. The cluster must have at least 10Gi free disk space, and should have at least 20Gi free disk space.
• A load balancer or ingress controller to load balance to the Paramify platform UI.
  • Must be configured for TLS communication to the backend.
  • If your load balancer or ingress controller requires explicit trust with the certificate presented by the backend service then you must upload a trusted certificate for the Paramify Web backend.
  • The Paramify platform UI must be accessed externally via port 443. Accessing Paramify over any other port is not supported.
  • Must support websockets.

Existing Cluster Installation​

Use these instructions to install Paramify into an existing cluster that has access to the internet.

Install the kubectl kots Plugin​

To install the latest version of the kots CLI to /usr/local/bin, run:

curl https://kots.io/install | bash

To install to a directory other than /usr/local/bin, run:

curl https://kots.io/install | REPL_INSTALL_PATH=/path/to/cli bash

To verify your installation, run:

kubectl kots --help

Install using sudo​

If you have sudo access to the directory where you want to install the kots CLI, you can set the REPL_USE_SUDO environment variable so that the installation script prompts you for your sudo password.

curl -L https://kots.io/install | REPL_USE_SUDO=y bash

Advanced Installation​

For an advanced installation of the kots cli, please refer to the official documentation.

Install the PPI Using the kubectl kots Plugin​

1. Choose the namespace that you will install the PPI and Paramify instance into. These instructions refer to this namespace as your-namespace.

2. First, run:

   kubectl kots install paramify --namespace your-namespace

   This automatically creates the namespace specified if it does not exist. If the KUBECONFIG environment variable is not set you will need to add --kubeconfig /path/to/kube/config to the install command above so that it can authenticate to your cluster.

3. After a few moments of initialization you are prompted for what you want the PPI Admin Console password to be.

   

4. When the install is complete it automatically proxies the PPI UI to http://localhost:8800. [insert image here] In a browser window, access the PPI UI at http://localhost:8800 to begin configuring the paramify platform.

5. Enter the password you specified during the install process.

   

6. Next, you are prompted to upload your .yaml license file. Select or drag the file, and then click "Upload license".

   

7. Finally, follow the steps for Configuring Paramify

8. Verify the preflight checks and proceed with the installation.

Configuring Paramify​

A Paramify installation can be customized to fit business needs via the provided configuration fields.

Single Sign-On (SSO)​

All authentication into Paramify is passwordless. The customer must configure SSO via an identity provider (Google, Okta, Microsoft, etc.) or authenticate via a magic link sent to their email.

SMTP​

An SMTP configuration is recommended for Paramify to function properly. Paramify sends emails for authentication via magic link, as well as emails for user invitations and notifications.

Cloud Storage​

A cloud storage provider is required for Paramify to function properly. The customer must provide a bucket and accompanying endpoints.

Example S3 Service Account Annotations

eks.amazonaws.com/role-arn: arn:aws:iam::$account_id:role/my-role

Example Azure Workload Identity Service Account Annotations

azure.workload.identity/client-id: <client_id>

CDN​

AWS Cloudflare is optional and supported.

Load Balancing​

Annotation Examples​

AWS load balancer:

# Use the following to assign an ACM cert to the loadbalancer
service.beta.kubernetes.io/aws-load-balancer-ssl-cert: arn:aws:acm:<region>:<account>:certificate/<id>
service.beta.kubernetes.io/aws-load-balancer-ssl-ports: http  # keep as "http"
service.beta.kubernetes.io/aws-load-balancer-backend-protocol: https  # keep as "https"

Azure load balancer:

# By default an external IP will be assigned, or use this for internal
service.beta.kubernetes.io/azure-load-balancer-internal: "true"

Database​

A Postgres database is required for Paramify to function properly. While Paramify supports an embedded database option, it is expected that the customer will provide an external database where snapshots and backups can be more effectively maintained.

If an embedded database is used then a basic automated backup option is available. See Embedded Postgres Backup for more details.