Deploying custom code to Data Grid

Guide
  • Red Hat Data Grid 8.3
  • Updated 08 February 2022
  • Published 02 December 2021

Topics included in this guide

Deploying custom code to Data Grid

Guide
Red Hat Data Grid 8.3
  • Updated 08 February 2022
  • Published 02 December 2021

Add custom code, such as scripts and event listeners, to your Data Grid clusters.

Before you can deploy custom code to Data Grid clusters, you need to make it available. To do this you can copy artifacts from a persistent volume (PV), download artifacts from an HTTP or FTP server, or use both methods.

Copying code artifacts to Data Grid clusters

Adding your artifacts to a persistent volume (PV) and then copy them to Data Grid pods.

This procedure explains how to use a temporary pod that mounts a persistent volume claim (PVC) that:

  • Lets you add code artifacts to the PV (perform a write operation).

  • Allows Data Grid pods to load code artifacts from the PV (perform a read operation).

To perform these read and write operations, you need certain PV access modes. However, support for different PVC access modes is platform dependent.

It is beyond the scope of this document to provide instructions for creating PVCs with different platforms. For simplicity, the following procedure shows a PVC with the ReadWriteMany access mode.

In some cases only the ReadOnlyMany or ReadWriteOnce access modes are available. You can use a combination of those access modes by reclaiming and reusing PVCs with the same spec.volumeName.

Using ReadWriteOnce access mode results in all Data Grid pods in a cluster being scheduled on the same OpenShift node.

Procedure

  1. Change to the namespace for your Data Grid cluster.

    $ oc project rhdg-namespace

  2. Create a PVC for your custom code artifacts, for example:

    apiVersion: v1
kind: PersistentVolumeClaim
metadata:
  name: datagrid-libs
spec:
  accessModes:
    - ReadWriteMany
  resources:
    requests:
      storage: 100Mi

  3. Apply your PVC.

    $ oc apply -f datagrid-libs.yaml

  4. Create a pod that mounts the PVC, for example:

    apiVersion: v1
kind: Pod
metadata:
  name: datagrid-libs-pod
spec:
  securityContext:
    fsGroup: 2000
  volumes:
    - name: lib-pv-storage
      persistentVolumeClaim:
        claimName: datagrid-libs
  containers:
    - name: lib-pv-container
      image: registry.redhat.io/datagrid/datagrid-8-rhel8:8.3
      volumeMounts:
        - mountPath: /tmp/libs
          name: lib-pv-storage

  5. Add the pod to the Data Grid namespace and wait for it to be ready.

    $ oc apply -f datagrid-libs-pod.yaml
$ oc wait --for=condition=ready --timeout=2m pod/datagrid-libs-pod

  6. Copy your code artifacts to the pod so that they are loaded into the PVC.

    For example to copy code artifacts from a local libs directory, do the following:

    $ oc cp --no-preserve=true libs datagrid-libs-pod:/tmp/

  7. Delete the pod.

    $ oc delete pod datagrid-libs-pod

    Specify the persistent volume with spec.dependencies.volumeClaimName in your Infinispan CR and then apply the changes.

    apiVersion: infinispan.org/v1
kind: Infinispan
metadata:
  name: infinispan
spec:
  replicas: 2
  dependencies:
    volumeClaimName: datagrid-libs
  service:
    type: DataGrid

If you update your custom code on the persistent volume, you must restart the Data Grid cluster so it can load the changes.

Additional resources

Downloading code artifacts

Add your artifacts to an HTTP or FTP server so that Data Grid Operator downloads them to the {lib_path} directory on each Data Grid node.

When downloading files, Data Grid Operator can automatically detect the file type. Data Grid Operator also extracts archived files, such as zip or tgz, to the filesystem after the download completes.

Each time Data Grid Operator creates a Data Grid node it downloads the artifacts to the node. The download also occurs when Data Grid Operator recreates pods after terminating them.

Prerequisites

  • Host your code artifacts on an HTTP or FTP server.

Procedure

  1. Add the spec.dependencies.artifacts field to your Infinispan CR.

    1. Specify the location of the file to download via HTTP or FTP as the value of the spec.dependencies.artifacts.url field.

    2. Optionally specify a checksum to verify the integrity of the download with the spec.dependencies.artifacts.hash field.

      The hash field requires a value is in the format of <algorithm>:<checksum> where <algorithm> is sha1|sha224|sha256|sha384|sha512|md5.

    3. Set the file type, if necessary, with the spec.dependencies.artifacts.type field.

      You should explicitly set the file type if it is not included in the URL or if the file type is actually different to the extension in the URL.

      If you set type: file, Data Grid Operator downloads the file as-is without extracting it to the filesystem.

      apiVersion: infinispan.org/v1
kind: Infinispan
metadata:
  name: infinispan
spec:
  replicas: 2
  dependencies:
    artifacts:
      - url: http://example.com:8080/path
        hash: sha256:596408848b56b5a23096baa110cd8b633c9a9aef2edd6b38943ade5b4edcd686
        type: zip
  service:
    type: DataGrid

  2. Apply the changes.

Legal Notices for Trademarks