This page describes the software, hardware, and networking requirements for the nodes where Shipa will be installed. Shipa can be installed on a single node or a highly available Kubernetes cluster.

Kubernetes Clusters

Today, Shipa supports Kubernetes clusters versions 1.10 and higher.

Tested and supported clusters are:

• On-premises clusters
• Minikube
• GKE (Google Kubernetes Engine)
• AKS (Azure Kubernetes Service)
• EKS (Amazon Elastic Kubernetes Service)
• OKE (Oracle Kubernetes Engine)
• RKE (Rancher Kubernetes Engine)

🚧

Single-zone clusters

Only single-zone clusters are supported on Shipa's free version.

Certain clusters, such as GKE, offer different versions of Linux and container runtimes. As described in the following section, Shipa supports the different Linux distributions, such as GKE's Container Optimized OS (cos).

Container Runtime Requirements

Shipa works with any modern Linux distribution.

For the container runtime, Shipa works with virtually any modern Docker version container runtime.

Containerd is also supported.

Shipa supports all 64-bit x86 operating systems.

Some distributions of Linux may have default firewall rules that block communication with Helm. The user should check the default firewall rules for Oracle Linux and open the required ports with firewalld if necessary.

CPU and Memory

Hardware requirements scale based on the size of your deployment that is controlled by Shipa. These resource consumptions are at the node level, nodes where Shipa will be installed.

Networking Requirements

During the install, Shipa creates and uses two public IPs for Load Balancers. These Load Balancer services expose some of the Shipa services, using the following ports:

By default, during the install, Shipa uses dynamic public IP set by the cloud provider. If preferred, static IP can be used instead by applying the optional parameter below when using Helm install:

--set service.nginx.loadBalancerIP=<Static IP>