Prerequisites

Install Required Tools

On the workstation you're installing from, you'll need:

Verify your kubectl version satisfies v1.28+:

kubectl version --client

Verify your helm version satisfies v3.12+:

helm version

Cloud Provider CLI (optional)

If you plan to provision a new cluster with the commands later on this page, you'll also need your cloud's CLI installed and authenticated:

If you're installing into a cluster that already exists and is reachable from kubectl, you can skip this section.

Cluster Requirements

BoltMCP runs on any conformant Kubernetes cluster (v1.28+). Pick the tier that matches your deployment:

A few notes on how these numbers are derived:

• Why 3 nodes for production. Rolling upgrades and node failures both take a node out of service. With 3 nodes, losing one during an upgrade still leaves a healthy majority, and pods on the failed node can reschedule onto the survivors. A 2-node cluster offers no margin.
• Why the per-node floor matters. Each node loses ~1 vCPU and ~1.5 GiB to the kubelet, OS, and platform agents (CNI, DNS, logging, metrics) before any BoltMCP pod runs. Two small nodes are not equivalent to one larger node.
• Postgres. The bundled Postgres is suitable for evaluation and small production deployments. For higher-traffic or compliance-sensitive workloads, point BoltMCP at a managed Postgres (Cloud SQL, RDS, Azure Database for PostgreSQL) and disable the bundled instance.
• Keycloak. Defaults to a single replica. For HA, scale to ≥2 replicas with a shared cache and sticky sessions at the ingress.
• Ingress controller. Not required at helm install time — pods will start and run without one. You'll install an ingress controller and configure TLS afterwards to expose the services externally (see the Ingress & TLS page).

Create a Cluster

gcloud container clusters create boltmcp-cluster \
  --zone europe-west2-a \
  --num-nodes 1 \
  --machine-type e2-standard-2

gcloud container clusters get-credentials boltmcp-cluster \
  --zone europe-west2-a

gcloud components install gke-gcloud-auth-plugin

export PATH="/opt/homebrew/share/google-cloud-sdk/bin:$PATH"

eksctl create cluster \
  --name boltmcp-cluster \
  --region eu-west-2 \
  --nodes 1 \
  --node-type t3.large

az provider register --namespace Microsoft.ContainerService

az provider show -n Microsoft.ContainerService --query registrationState -o tsv

az group create --name boltmcp-rg --location westeurope

az aks create \
  --resource-group boltmcp-rg \
  --name boltmcp-cluster \
  --node-count 1 \
  --node-vm-size Standard_D2s_v3 \
  --generate-ssh-keys

az aks get-credentials \
  --resource-group boltmcp-rg \
  --name boltmcp-cluster

Verify Cluster Access

Confirm kubectl is pointing at the cluster you intend to install into:

kubectl config current-context

If it's not the right one, list the available contexts and switch:

kubectl config get-contexts

kubectl config use-context <name>

Finally, verify the cluster is ready:

kubectl get nodes

All nodes should show Ready status.