[CIE-5419] CDH migration from appliance to k8s edge

content/edge-kubernetes/manage-edge-bundle/migrating-edge-appliance.md

@@ -197,6 +197,20 @@ Perform the following steps as a root user on your Edge appliance.
 
 4. Tar the MongoDB data and data lake contents from {{< product-c8y-iot >}} DataHub if present, into */opt/edge-appliance-backup.tar* :
 
+   1. Migrate HSQL db.
+
+      ```shell
+      pip install https://download.cumulocity.com/Cumulocity-Edge/Installer/2025/cdh_migration-2025-py3-none-any.whl \
+      && cdh-migration
+      ```
+      wait until you see an out put `INFO - Data export complete: database_export.sql`
+   2.
+      ```shell
+      tar -zcf /opt/edge-appliance-backup.tar /opt/appliance-edgedb-backup /opt/softwareag /opt/mongodb/cdh-dremio/distributed-storage /opt/mongodb/cdh-master/datalake /home/admin/database_export.sql
+      ```
+
+   If DataHub is not present:
+
    ```shell
    tar -zcf /opt/edge-appliance-backup.tar /opt/appliance-edgedb-backup /opt/softwareag
    ```
@@ -247,8 +261,78 @@ After installing and configuring Edge 2025, proceed to migrate the data backed u
    ```shell
    rm -rf /opt/appliance-edgedb-backup /opt/edge-appliance-backup.tar
    ```
+### Step 5 - Restore DataHub
+1. Stop Edge operator:
+   ```shell
+   kubectl scale deployment c8yedge-operator-controller-manager -n c8yedge --replicas=0
+   ```
+
+2. Set tenant option `CDH_PASSWORD_SECRET` on edge tenant:
+   1. Get the secret from backup:
+      ```shell
+      grep CDH_PASSWORD_SECRET /opt/softwareag/cdh-console/conf/cdh-console-env
+      ```
+   2. Set tenant option on edge:
+      ```shell
+      curl --location 'https://<EDGE_HOST_IP>/tenant/options/' \
+      -H 'Content-Type: application/vnd.com.nsn.cumulocity.option+json' \
+      -H 'Accept: application/vnd.com.nsn.cumulocity.option+json' \
+      -u "management/${MANAGEMENT_ADMIN_USER}:${MANAGEMENT_ADMIN_PASSWORD}" \
+      --data '{
+         "category": "datahub",
+         "key": "credentials.CDH_PASSWORD_SECRET",
+         "value": "<PASSWORD_SECRET_FROM_PREVIOUS_STEP>"
+      }'
+      ```
+
+3. Migrate dremio:
+   1. Redeploy dremio in maintanace mode for migration, fetch dremio helm chart from installartifacts:
+      ```shell
+       kubectl get pv -A -o yaml | grep -A1 installartifacts-backing | awk '/path:/ {print $2}'
+      ```
+   2. Backup values.yaml:
+      ```shell
+      helm get values dremio -n c8yedge --all -o yaml > dremio_values.yaml
+      ```
+   3. Set dremio to maintanance:
+      ```shell
+      helm upgrade dremio ./dremio-11.0.648.tgz   -n c8yedge   -f dremio_values.yaml   --set DremioAdmin=true   --wait
+      ```
+   4. Restore RocksDB from backup:
+      ```shell
+      kubectl exec -n c8yedge dremio-admin -- rm -rf /opt/dremio/data/db && kubectl cp /opt/mongodb/cdh-master/data/db dremio-admin:/opt/dremio/data/ -n c8yedge
+      ```
+   5. Restore datalake contents:
+      ```shell
+      rm -rf /datahub/distributedStorage/* /datahub/datalake/* && cp -a /opt/mongodb/cdh-dremio/distributed-storage/. /datahub/distributedStorage/ && cp -a /opt/mongodb/cdh-master/datalake/. /datahub/datalake/
+      ```
+   6. Exit from maintaince mode:
+      ```shell
+      helm upgrade dremio ./dremio-11.0.648.tgz   -n c8yedge   -f dremio_values.yaml   --set DremioAdmin=false   --wait
+      ```
+
+4. Restore DataHub:
+   ```shell
+   kubectl cp /home/admin/database_export.sql -n c8yedge datahub-mysql-0:/tmp/database_export.sql && \
+   kubectl exec -i -n c8yedge datahub-mysql-0 -- sh -c 'mysql -u root -p"$MYSQL_ROOT_PASSWORD" CDH_edge < /tmp/database_export.sql'
+   ```
+
+5. Configure dremio:
+   Post migration, dremio will have legacy configuration of c8y_source from appliance, this needs to be updated for dremio to communicate with mongo. In this step we will update the `c8y_source` configuration:
+      1. Set support property “store.plugin.keep_metadata_on_replace“ = true in Dremio (goto Settings > Support)
+      2. Configure settings of “c8y-source” as required for edge on K8s:
+         a. Configure MongoDB host name and credentials 
+      3. Set support property “store.plugin.keep_metadata_on_replace“ = false in Dremio
+   ```shell
+   curl -sfL {{< link-c8y-doc-baseurl >}}files/edge-k8s/dremio_source_configuration.sh -O && bash ./dremio_source_configuration.sh
+   ```
+
+6. Start Edge operator:
+   ```shell
+   kubectl scale deployment c8yedge-operator-controller-manager -n c8yedge --replicas=1
+   ```
 
-### Step 5 - Configuring Edge 2025 post migration
+### Step 6 - Configuring Edge 2025 post migration
 After successfully migrating your data to Edge 2025, you'll need to configure it to match your previous Edge Appliance VM setup. Here's what you need to do:
 
 #### What's already available?

static/files/edge-k8s/dremio_source_configuration.sh

@@ -0,0 +1,145 @@
+#!/bin/bash
+set -euo pipefail
+trap 'echo "Error on line $LINENO"; exit 1' ERR
+
+# ------- Configuration (no manual inputs) -------
+NAMESPACE="c8yedge"
+SOURCE_NAME="c8y_source"
+MONGO_SECRET="internal-edge-db-users"
+DREMIO_SECRET="dremio-admin-credentials"
+MONGO_SVC="edge-db-rs0"
+DREMIO_SVC="cumulocity-ontoplb"
+DREMIO_PORT=9047
+
+# ------- Fetch Dremio admin credentials -------
+DREMIO_USER=$(kubectl get secret -n "$NAMESPACE" "$DREMIO_SECRET" -o jsonpath='{.data.DREMIO_ADMIN_USER}' | base64 -d)
+DREMIO_PASS=$(kubectl get secret -n "$NAMESPACE" "$DREMIO_SECRET" -o jsonpath='{.data.DREMIO_ADMIN_PASSWORD}' | base64 -d)
+
+# ------- Resolve Dremio endpoint (LoadBalancer IP or ClusterIP fallback) -------
+DREMIO_IP=$(kubectl get svc -n "$NAMESPACE" "$DREMIO_SVC" -o jsonpath='{.status.loadBalancer.ingress[0].ip}' 2>/dev/null || true)
+if [[ -z "$DREMIO_IP" || "$DREMIO_IP" == "null" ]]; then
+  DREMIO_IP=$(kubectl get svc -n "$NAMESPACE" "$DREMIO_SVC" -o jsonpath='{.spec.clusterIP}')
+fi
+if [[ -z "$DREMIO_IP" ]]; then
+  echo "Failed to locate Dremio service IP for $DREMIO_SVC in namespace $NAMESPACE"
+  exit 1
+fi
+DREMIO_URL="https://${DREMIO_IP}:${DREMIO_PORT}"
+
+# ------- Login to Dremio and build Authorization header -------
+LOGIN_JSON=$(curl -sk -X POST "$DREMIO_URL/apiv2/login" \
+  -H "Content-Type: application/json" \
+  -d "{\"userName\":\"$DREMIO_USER\",\"password\":\"$DREMIO_PASS\"}")
+
+AUTH_TOKEN=$(echo "$LOGIN_JSON" | jq -r '.token')
+if [[ -z "$AUTH_TOKEN" || "$AUTH_TOKEN" == "null" ]]; then
+  echo "Failed to obtain Dremio auth token. Response:"
+  echo "$LOGIN_JSON" | jq .
+  exit 1
+fi
+AUTH_HEADER="_dremio${AUTH_TOKEN}"
+
+# ------- Fetch Mongo credentials -------
+MONGO_USER=$(kubectl get secret -n "$NAMESPACE" "$MONGO_SECRET" -o jsonpath='{.data.MONGODB_DATABASE_ADMIN_USER}' | base64 -d)
+MONGO_PASS=$(kubectl get secret -n "$NAMESPACE" "$MONGO_SECRET" -o jsonpath='{.data.MONGODB_DATABASE_ADMIN_PASSWORD}' | base64 -d)
+
+# ------- Resolve Mongo service host and port -------
+MONGO_HOSTNAME=$(kubectl get svc -n "$NAMESPACE" "$MONGO_SVC" -o jsonpath='{.metadata.name}' 2>/dev/null || true)
+if [[ -z "$MONGO_HOSTNAME" ]]; then
+  echo "Failed to find MongoDB service $MONGO_SVC in namespace $NAMESPACE"
+  exit 1
+fi
+MONGO_PORT=$(kubectl get svc -n "$NAMESPACE" "$MONGO_SVC" -o jsonpath='{.spec.ports[0].port}')
+MONGO_FQDN="${MONGO_HOSTNAME}.${NAMESPACE}.svc.cluster.local"
+
+# ------- Fetch existing source JSON from Dremio -------
+SOURCE_JSON=$(curl -sk -H "Authorization: ${AUTH_HEADER}" "$DREMIO_URL/apiv2/source/$SOURCE_NAME")
+SOURCE_ID=$(echo "$SOURCE_JSON" | jq -r '.id // empty')
+SOURCE_TAG=$(echo "$SOURCE_JSON" | jq -r '.tag // empty')
+
+if [[ -z "$SOURCE_ID" ]]; then
+  echo "Source '$SOURCE_NAME' not found in Dremio at $DREMIO_URL"
+  exit 1
+fi
+
+# ------- Enable keep_metadata_on_replace (true) -------
+curl -sk -X PUT "$DREMIO_URL/apiv2/settings/store.plugin.keep_metadata_on_replace" \
+  -H "Authorization: ${AUTH_HEADER}" \
+  -H "Content-Type: application/json" \
+  --data '{"type":"BOOLEAN","id":"store.plugin.keep_metadata_on_replace","value":true}' >/dev/null
+
+# ------- Build minimal update payload based on requested layout -------
+# Layout:
+# {
+#   "config": { hostList:[{hostname,port}], useSsl:true, authenticationType, authDatabase, ... }
+# }
+UPDATE_PAYLOAD=$(jq -n \
+  --arg id "$SOURCE_ID" \
+  --arg tag "$SOURCE_TAG" \
+  --arg name "$SOURCE_NAME" \
+  --arg type "MONGO" \
+  --arg host "$MONGO_FQDN" \
+  --argjson port "$MONGO_PORT" \
+  --arg user "$MONGO_USER" \
+  --arg pass "$MONGO_PASS" \
+  '{
+    id: $id,
+    tag: $tag,
+    type: $type,
+    name: $name,
+    config: {
+      hostList: [{ hostname: $host, port: $port }],
+      useSsl: true,
+      authenticationType: "MASTER",
+      username: $user,
+      password: $pass,
+      authDatabase: "admin",
+      authenticationTimeoutMillis: 2000,
+      secondaryReadsOnly: false,
+      subpartitionSize: 0,
+      sampleSize: 4095,
+      sampleMethod: "FIRST",
+      propertyList: [{ name: "maxPoolSize", value: "100" }],
+      useCaseInsensitiveFieldNames: false
+    }
+  }')
+
+# ------- PUT update to Dremio -------
+UPDATE_RESPONSE=$(curl -sk -X PUT "$DREMIO_URL/apiv2/source/$SOURCE_NAME/?nocache=$(date +%s)" \
+  -H "Authorization: ${AUTH_HEADER}" \
+  -H "Content-Type: application/json" \
+  --data "$UPDATE_PAYLOAD")
+
+# If Dremio returned an errorMessage, show and exit
+if echo "$UPDATE_RESPONSE" | jq -e 'has("errorMessage")' >/dev/null 2>&1; then
+  echo "Dremio returned error while updating source:"
+  echo "$UPDATE_RESPONSE" | jq .
+  # Attempt to restore the keep_metadata flag to false before exiting
+  curl -sk -X PUT "$DREMIO_URL/apiv2/settings/store.plugin.keep_metadata_on_replace" \
+    -H "Authorization: ${AUTH_HEADER}" \
+    -H "Content-Type: application/json" \
+    --data '{"type":"BOOLEAN","id":"store.plugin.keep_metadata_on_replace","value":false}' >/dev/null || true
+  exit 1
+fi
+
+# ------- Restore keep_metadata_on_replace (false) -------
+curl -sk -X PUT "$DREMIO_URL/apiv2/settings/store.plugin.keep_metadata_on_replace" \
+  -H "Authorization: ${AUTH_HEADER}" \
+  -H "Content-Type: application/json" \
+  --data '{"type":"BOOLEAN","id":"store.plugin.keep_metadata_on_replace","value":false}' >/dev/null
+
+# ------- Verify connection/status -------
+VERIFY_JSON=$(curl -sk -H "Authorization: ${AUTH_HEADER}" "$DREMIO_URL/apiv2/source/$SOURCE_NAME")
+STATUS=$(echo "$VERIFY_JSON" | jq -r '.state.status // empty')
+MSG=$(echo "$VERIFY_JSON" | jq -r '.state.messages[0].message // empty')
+
+if [[ "$STATUS" == "good" ]]; then
+  echo "SUCCESS: Dremio source '$SOURCE_NAME' updated and MongoDB is connected."
+  echo "Message: $MSG"
+  exit 0
+else
+  echo "WARNING: Dremio source updated but status is not 'good'."
+  echo "Status: $STATUS"
+  echo "Message: $MSG"
+  exit 2
+fi