Resources:

In this sample deployment, we increased the minimum cpu and memory request to 2cpu and 4G memory with an upper limit of 6cpu and 8G memory for each artifactory node. We also increased Jvm memory to 4g xms and 6g xmx. For detailed system resources recommendation, please refer to here .

The default properties that artifactor use may not be fully utilizing the available system resources. If you find artifactory underperformed with enough physical resources please check out the tuning section in the glossary.

Scalability is controlled by artifactory.node.replicaCount. We kept the default value 2, which results in a 3 node cluster.

You can test the properties with the following dry-run installation:

$ helm install artifactory-ha \
    --set artifactory.primary.resources.requests.cpu="2" \
    --set artifactory.primary.resources.limits.cpu="6" \
    --set artifactory.primary.resources.requests.memory="4Gi" \
    --set artifactory.primary.resources.limits.memory="8Gi" \
    --set artifactory.primary.javaOpts.xms="4g" \
    --set artifactory.primary.javaOpts.xmx="6g" \
    --set artifactory.node.resources.requests.cpu="2" \
    --set artifactory.node.resources.limits.cpu="6" \
    --set artifactory.node.resources.requests.memory="4Gi" \
    --set artifactory.node.resources.limits.memory="8Gi" \
    --set artifactory.node.javaOpts.xms="4g" \
    --set artifactory.node.javaOpts.xmx="6g" \
    --set artifactory.node.replicaCount=2\
    --namespace artifactory-ha jfrog/artifactory-ha --version 107.21.12 --dry-run

We will add the properties above to a values.yaml.

artifactory:
  primary:
    resources:
      requests:
        memory: "4Gi"
        cpu: "2"
      limits:
        memory: "8Gi"
        cpu: "6"
    javaOpts:
      xms: "4g"
      xmx: "6g"
  node:
    replicaCount: 2
    resources:
      requests:
        memory: "4Gi"
        cpu: "2"
      limits:
        memory: "8Gi"
        cpu: "6"
    javaOpts:
      xms: "4g"
      xmx: "6g"