DX-41560: Change Zookeeper to use official Zookeeper images

- Changes Dremio v2 Helm chart to utilize the official Zookeeper image.
  ** Note: Using the old previous version of Zookeeper with this chart
  change is not supported. **
- Updates to allow 3.8.0+ Zookeeper versions to be used.
- Adds a check to fail when using an older, unsupported ZK image.
- For the latest Zookeeper image available, see:
  https://hub.docker.com/_/zookeeper

Change-Id: I4a9f26c83036ec17ae47992553edcec224816d51

charts/dremio_v2/README.md

@@ -107,3 +107,14 @@ $ kubectl get services dremio-client
 
    If you want to change the expose port on the load balancer, change the value of the setting `coordinator.flight.port` in the file `values.local.yaml`.
 * If the value in the TYPE column of the output is `NodePort`, access Dremio using Flight through http://localhost:31357.
+
+## Upgrading Zookeeper
+
+To upgrade to the official Zookeeper image set the following tags in `values.local.yaml` file.
+
+```
+zookeeper:
+  image: zookeeper
+  imageTag: 3.8.0
+```
+

charts/dremio_v2/docs/Values-Reference.md

@@ -1462,15 +1462,15 @@ When setting this property, the load balancer attempts to use the provided IP ad
 
 Type: String
 
-By default, the value is set to `k8s.gcr.io/kubernetes-zookeeper`.
+By default, the value is set to `zookeeper`.
 
 The `image` refers to the location to retrieve the specific container image for Zookeeper. In some cases, the `zookeeper.image` value may vary in corporate environments where there may be a private container registry that is used.
 
 #### `zookeeper.imageTag`
 
 Type: String
 
-By default, the value is set to `1.0-3.4.10`.
+By default, the value is set to `3.8.0`.
 
 The version of Zookeeper set has been validated by Dremio to work with the Dremio software. Changing this version is not recommended unless the tag is different due to a private container registry name difference.
 

charts/dremio_v2/templates/zookeeper.yaml

@@ -78,6 +78,36 @@ spec:
       containers:
       - name: kubernetes-zookeeper
         imagePullPolicy: Always
+        command:
+        - bash
+        - -ec
+        - |
+            # obtain ZOO_MY_ID based on POD hostname
+            export HOST_NUMBER="${HOSTNAME##*-}"
+            export ZOO_MY_ID="$((HOST_NUMBER + 1))"
+            [ -z "${ZOO_MY_ID}" ] && echo "Failed to get index from hostname $HOSTNAME" && exit 1
+            echo "ZOO_MY_ID=${ZOO_MY_ID}"
+
+            # construct ZOO_SERVERS based on ensemble count
+            DOMAIN="$(hostname -d)"
+            SERVERS={{ $.Values.zookeeper.count }}
+            for (( i=${SERVERS},j=i-1; i>=1; i--,j-- )); do ZOO_SERVERS="server.${i}=zk-${j}.${DOMAIN}:2888:3888;2181 ${ZOO_SERVERS}"; done
+            echo "ZOO_SERVERS=${ZOO_SERVERS}"
+            export ZOO_SERVERS
+
+            /docker-entrypoint.sh
+            zkServer.sh start-foreground
+        env:
+        - name: JVMFLAGS
+          value: "-Xmx{{ template "dremio.zookeeper.memory" $ }}m"
+        - name: ZOO_STANDALONE_ENABLED
+          value: "false"
+        - name: ZOO_4LW_COMMANDS_WHITELIST
+          value: "ruok"
+        - name: ZOO_ADMINSERVER_ENABLED
+          value: "false"
+        - name: ZOO_AUTOPURGE_PURGEINTERVAL
+          value: "12"
         image: "{{ $.Values.zookeeper.image }}:{{ $.Values.zookeeper.imageTag }}"
         resources:
           requests:
@@ -90,40 +120,19 @@ spec:
           name: server
         - containerPort: 3888
           name: leader-election
-        command:
-        - sh
-        - -c
-        - "start-zookeeper \
-          --servers={{ $.Values.zookeeper.count }} \
-          --data_dir=/var/lib/zookeeper/data \
-          --data_log_dir=/var/lib/zookeeper/data/log \
-          --conf_dir=/opt/zookeeper/conf \
-          --client_port=2181 \
-          --election_port=3888 \
-          --server_port=2888 \
-          --tick_time=2000 \
-          --init_limit=10 \
-          --sync_limit=5 \
-          --heap={{- template "dremio.zookeeper.memory" $ -}}M \
-          --max_client_cnxns=60 \
-          --snap_retain_count=3 \
-          --purge_interval=12 \
-          --max_session_timeout=40000 \
-          --min_session_timeout=4000 \
-          --log_level=INFO"
         readinessProbe:
           exec:
-            command: ["sh", "-c", "zookeeper-ready 2181"]
+            command: ["/bin/bash", "-c", "[ \"$(echo ruok | nc 127.0.0.1 2181)\" == \"imok\" ]" ]
           initialDelaySeconds: 10
           timeoutSeconds: 5
         livenessProbe:
           exec:
-            command: ["sh", "-c", "zookeeper-ready 2181"]
+            command: ["/bin/bash", "-c", "[ \"$(echo ruok | nc 127.0.0.1 2181)\" == \"imok\" ]" ]
           initialDelaySeconds: 10
           timeoutSeconds: 5
         volumeMounts:
         - name: datadir
-          mountPath: /var/lib/zookeeper
+          mountPath: /data
       securityContext:
         runAsUser: 1000
         fsGroup: 1000
@@ -137,4 +146,10 @@ spec:
       resources:
         requests:
           storage: {{ $.Values.zookeeper.volumeSize }}
-{{- end -}}
+
+# use the official zookeeper kubernetes image
+  {{- if (eq $.Values.zookeeper.image "k8s.gcr.io/kubernetes-zookeeper") }}
+    {{- fail "\n\n** ERROR: Zookeeper image 'k8s.gcr.io/kubernetes-zookeeper' is no longer supported. Please see 'Upgrading Zookeeper' section in README.md. **\n" -}}
+  {{- end }}
+
+{{- end -}}

charts/dremio_v2/values.yaml

@@ -254,8 +254,8 @@ executor:
 # Zookeeper
 zookeeper:
   # The Zookeeper image used in the cluster.
-  image: k8s.gcr.io/kubernetes-zookeeper
-  imageTag: 1.0-3.4.10
+  image: zookeeper
+  imageTag: 3.8.0
 
   # CPU & Memory
   # Memory allocated to each zookeeper, expressed in MB.
@@ -474,4 +474,4 @@ service:
 # here.  For more information, see
 # https://kubernetes.io/docs/concepts/containers/images/#specifying-imagepullsecrets-on-a-pod
 # imagePullSecrets:
-#  - secretname
+#  - secretname