Add support for JDBCConfig family of active clustering

geoserver/latest/JDBCConfig_readme.md

@@ -0,0 +1,83 @@
+# Instructions 
+
+Instructions to install GS with the JDBCConfig family of plugins:
+
+1. Install GS with plugins using our helm chart with following custom_values.yaml:
+
+(Adjust the version in URLs to match the GS version image that you use)
+
+```yaml
+geoserver:
+    plugins: "https://build.geoserver.org/geoserver/2.26.x/ext-latest/geoserver-2.26-SNAPSHOT-wps-plugin.zip https://build.geoserver.org/geoserver/2.26.x/ext-latest/geoserver-2.26-SNAPSHOT-wps-cluster-hazelcast-plugin.zip https://build.geoserver.org/geoserver/2.26.x/community-latest/geoserver-2.26-SNAPSHOT-jdbcconfig-plugin.zip https://build.geoserver.org/geoserver/2.26.x/community-latest/geoserver-2.26-SNAPSHOT-jdbcstore-plugin.zip https://build.geoserver.org/geoserver/2.26.x/community-latest/geoserver-2.26-SNAPSHOT-hz-cluster-plugin.zip"
+```
+
+```sh
+helm install geoserver . --values custom_values.yaml
+```
+
+2. Edit the following configmaps and ensure correct DB string and credentials in all of them:
+
+```sh
+kubectl edit configmap/cm-jdbc-init-import
+kubectl edit configmap/cm-jdbc-enabled
+kubectl edit configmap/cm-jdbcstore-init-import
+kubectl edit configmap/cm-jdbcstore-enabled
+```
+
+Also edit `kubectl edit configmap/hazelcast-xml` and adjust the namespace of the FQDN string `standalone-hz.default.svc.cluster.local` as necessary. (For example, If you installed this helm chart in a different namespace)
+
+3. Patch the geoserver statefulset to mount the "cm-jdbc-init-import" and "cm-jdbcstore-init-import" configmap and replicaCount is set to 1:
+
+```sh
+kubectl patch statefulset geoserver --type='json' -p='[
+  {
+    "op": "replace",
+    "path": "/spec/template/spec/containers/0/volumeMounts/0/name",
+    "value": "cm-jdbc-init-import"
+  },
+  {
+    "op": "replace",
+    "path": "/spec/template/spec/containers/0/volumeMounts/1/name",
+    "value": "cm-jdbcstore-init-import"
+  },
+  {
+    "op": "replace",
+    "path": "/spec/replicas",
+    "value": 1
+  }
+]'
+```
+
+4. At this point, geoserver pod will restart and the jdbcstore import to the DB will take place. Monitor the logs and wait for the process to complete. You can check the DB and see if "objects" and "resources" relations are created in the public namespace.
+
+```sh
+kubectl logs -f statefulsets/geoserver
+```
+
+5. Patch the geoserver statefulset to use the "cm-jdbc-enabled" and "cm-jdbcstore-enabled" configmap and to your desired replica count:
+
+```sh
+kubectl patch statefulset geoserver --type='json' -p='[
+  {
+    "op": "replace",
+    "path": "/spec/template/spec/containers/0/volumeMounts/0/name",
+    "value": "cm-jdbc-enabled"
+  },
+  {
+    "op": "replace",
+    "path": "/spec/template/spec/containers/0/volumeMounts/1/name",
+    "value": "cm-jdbcstore-enabled"
+  },
+    {
+    "op": "replace",
+    "path": "/spec/replicas",
+    "value": 2
+  }
+]'
+```
+
+6. Ensure geoserver is restarted
+
+---
+
+NB: Repeat step 2 and 5 if helm chart is upgraded.

geoserver/latest/templates/jdbcconfig/cm-cluster_properties.yaml

@@ -0,0 +1,18 @@
+apiVersion: v1
+kind: ConfigMap
+metadata:
+  name: cluster-properties
+data:
+  cluster.properties: |
+    enabled = true
+    sync_method = event
+    sync_delay = 5
+    session_sharing = true
+    session_sticky = false
+    acktimeout = 2000
+    #Disabling the hazelcast shutdown hook
+    shutdown_hook_enable = false
+    #Setting the hazelcast logging type to log4j
+    logging_type = "log4j"
+    logger.com-hazelcast.name = com.hazelcast
+    logger.com-hazelcast.level = ERROR