Openshift only allows statefulsets / pods to run in specific user and group id ranges. Artifactory currently uses users outside of this allowed range. For this reason the service account for the operator in the jfrog-artifactory namespace must be granted anyuid privileges.

Ex: oc adm policy add-scc-to-user anyuid system:serviceaccount:jfrog-artifactory:artifactory-ha-operator

Where anyuid is the Security context constraint being applied to the service account artifactory-ha-operator in namespace jfrog-artifactory.

If you run setup.sh these will be created on the cluster your kubectl or oc program is connected to.

Security Context Constraints - Hostpath

Openshift does not have the hostpath plugin enabled by default.

A security context constraint has been created for hostpath in deploy/hostpathscc.yaml

You can apply the security context constraint and hostpath plugin patch via these commands:

oc apply -f deploy/hostpathscc.yaml oc patch securitycontextconstraints.security.openshift.io/hostpath --type=merge --patch='{"allowHostDirVolumePlugin": true}' oc adm policy add-scc-to-user hostpath system:serviceaccount:jfrog-artifactory:artifactory-ha-operator

Or if you run setup.sh these will already be done.

Persistent Volumes

Artifactory HA nodes by default request persistent volume claims 200 Gbs in size.

If your cluster does not already have existing persistent volumes that are 200Gi you will need to create new persistent volumes that are large enough to bound the claims to.

Example persistent volumes can be found at:

helm-charts/openshift-artifactory-ha/pv-examples

If you create the five folders on each node:

/mnt/pv-data/pv0001-large /mnt/pv-data/pv0002-large /mnt/pv-data/pv0003-large /mnt/pv-data/pv0004-large /mnt/pv-data/pv0005-large

You can then apply the example persistent volumes to your cluster with:

oc apply -f helm-charts/openshift-artifactory-ha/pv-examples

Installation types

OLM Catalog

To install via the OLM catalog download the operator from the Operator hub and install it via the Openshift console GUI

To test OLM catalog installs you will need to deploy the lastest ClusterServiceVersion found at: deploy/olm-catalog/artifactory-ha-operator/X.X.X/artifactory-ha-operator.vX.X.X.clusterserviceversion.yaml

This will install the operator into whatever cluster your kubectl or oc program is currently logged into.

Please refer to Local Testing section below for full instructions.

Operator YAML

To install the operator via the Operator YAML follow the Local Testing tests.

Instead of running operator-sdk up local for the last step run:

oc apply -f deploy/operator.yaml

This will deploy the operator into the cluster.

Local Testing

Please refer to cluster setup. Ensure all steps have been completed prior to local testing against code ready containers.

Follow these steps:

Install code ready containers if you do not already have it installed.

Run your cluster with 2 cpus and 8192 MBs of memory at a minimum to support HA:

crc start -c 2 -m 8192

Recommended settings:

crc start -c 4 -m 16384

Create file: JFrog-Cloud-Installers/Openshift4/artifactory.cluster.license Paste your license keys into this file for HA configuration of multiple nodes. Note: License keys must be separated by two new lines.

Run: setup.sh

Operator-sdk local

Run: operator-sdk up local