alachisoft/ncache-helm repository overview

⁠NCache Helm Chart

Deploys an NCache⁠ distributed cache cluster on Kubernetes (on-premises, AKS, EKS). The chart creates the NCache server pods, the client and management gateways, and the caches you enable, and handles licensing, TLS, scaling and image upgrades through helm install / helm upgrade.

To see every configurable parameter and its default:

helm show values oci://registry-1.docker.io/alachisoft/ncache-helm --version 5.3.8

⁠Install (minimum)

Only the installation key is required. deployment.platformcan be used to match your cluster (on-prem, AKS, EKS). But is option paramter and its default value is on-prem

helm install ncache oci://registry-1.docker.io/alachisoft/ncache-helm --version 5.3.8 \
  --set deployment.key="INSTALLATION-KEY"

This starts a 2-server cluster with the cache defaultCache (Partition-Replica, 1 GB) and runs on an evaluation license.

Check the deployment:

kubectl get pods 
kubectl get services 

⁠Register NCache with a License Key

Activate a purchased license at install time:

helm install ncache oci://registry-1.docker.io/alachisoft/ncache-helm --version 5.3.8 \
  --set deployment.key="INSTALLATION-KEY" \
  --set deployment.platform=on-prem \
  --set license.licenseKey="XXXXXXXX-XXXXXX-XXXXXXXX" \
  --set license.envName="Prod"

Or on a running release:

helm upgrade ncache oci://registry-1.docker.io/alachisoft/ncache-helm --version 5.3.8 \
  --reuse-values \
  --set license.licenseKey="XXXXXXXX-XXXXXX-XXXXXXXX" \
  --set license.envName="Prod"

⁠Caches

The chart has four cache slots. caches[0] (defaultCache) is enabled by default; enable the others in order (caches[1] before caches[2]). Disabling a cache removes it and its data.

Create a cache at install time with your own settings:

helm install ncache oci://registry-1.docker.io/alachisoft/ncache-helm --version 5.3.8 \
  --set deployment.key="INSTALLATION-KEY" \
  --set 'caches[1].enabled=true' \
  --set 'caches[2].enabled=true' \
  --set 'caches[3].enabled=true'

Add a cache to a running release:

helm upgrade ncache oci://registry-1.docker.io/alachisoft/ncache-helm --version 5.3.8 \
  --reuse-values \
  --set 'caches[1].enabled=true' \
  --set 'caches[2].enabled=true' \
  --set 'caches[3].enabled=true'

⁠TLS

TLS can be set at install time or switched on and off on a running release; the servers are not replaced, only the caches are restarted one server at a time.

helm install ncache oci://registry-1.docker.io/alachisoft/ncache-helm --version 5.3.8 \
  --set deployment.key="INSTALLATION-KEY" \
  --set tls.enabled=true

Enable or disable on a running release:

helm upgrade ncache oci://registry-1.docker.io/alachisoft/ncache-helm --version 5.3.8 \
  --reuse-values --set tls.enabled=true

⁠Gateway with a fixed public IP

The chart publishes two LoadBalancer services:

By default the load balancer assigns the address. To give the management gateway a fixed address, set gateway.mgmtCenterEndpoint:

helm install ncache oci://registry-1.docker.io/alachisoft/ncache-helm --version 5.3.8 \
  --set deployment.key="INSTALLATION-KEY" \
  --set gateway.mgmtCenterEndpoint="20.10.30.40"

Enable or disable on a running release:

helm upgrade ncache oci://registry-1.docker.io/alachisoft/ncache-helm --version 5.3.8 \
  --reuse-values -gateway.mgmtCenterEndpoint="20.10.30.40"

• on-prem: use an unused IP from the LoadBalancer e.g MetalLB pool (not a node IP).

Find the gateway addresses:

kubectl get service ncache-mgmt-gateway ncache-client-gateway 

Then open Management Center at http://<EXTERNAL-IP of ncache-mgmt-gateway>:9801 and point clients at <EXTERNAL-IP of ncache-client-gateway>:9800.

⁠Scaling with helm upgrade

Change the number of NCache servers on a running release. Scale-up adds the pods and joins them to every cache; scale-down removes one server at a time so each partition's replica is rebuilt before the next server leaves.

Scale up to 4 servers:

helm upgrade ncache oci://registry-1.docker.io/alachisoft/ncache-helm --version 5.3.8 \
  --reuse-values \
  --set deployment.servers=4

Scale back down to 2 servers:

helm upgrade ncache oci://registry-1.docker.io/alachisoft/ncache-helm --version 5.3.8 \
  --reuse-values \
  --set deployment.servers=2

Watch the change:

kubectl get pods  -w

Always pass --reuse-values to helm upgrade; without it Helm resets every setting you do not repeat to its default (enabled caches would be removed). deployment.name, deployment.podAntiAffinity and deployment.platform are fixed at install time and an upgrade that changes them is rejected.

⁠Uninstall

helm uninstall ncache

⁠Support

• Documentation: https://www.alachisoft.com/resources/docs/⁠
• Support: [email protected]⁠