Post start

dockerization/Dockerfile

@@ -32,14 +32,18 @@ RUN chmod +x /opt/drill/bin/*
 RUN chown drill:drill /tmp/drill
 RUN chmod 700 /tmp/drill
 
-RUN paxctl -c /usr/lib/jvm/java-8-oracle/jre/bin/java && paxctl -m /usr/lib/jvm/java-8-oracle/jre/bin/java 
+RUN paxctl -c /usr/lib/jvm/java-8-openjdk-amd64/jre/bin/java && paxctl -m /usr/lib/jvm/java-8-openjdk-amd64/jre/bin/java
 
 COPY files/drill_ddl.sql /
 COPY files/test.sql /tmp/
+COPY files/post-start.sh /
 COPY files/docker-entrypoint.sh /
+COPY files/wait-for.sh /
 
 COPY files/version.sh /version.sh
 RUN ln -s /version.sh /usr/bin/version
+RUN ln -s /post-start.sh /usr/bin/post-start
+RUN ln -s /wait-for.sh /usr/bin/wait-for
 
 USER drill
 CMD ["/docker-entrypoint.sh"]

dockerization/files/post-start.sh

@@ -0,0 +1,42 @@
+#!/bin/bash
+
+readonly HOSTNAME=$(hostname)
+readonly PORT=8047
+
+echo "INFO: waiting for Drill Java initialization."
+echo "INFO: Checking if Drill is up ..."
+wait-for "$HOSTNAME" "$PORT" || exit 1
+
+chown drill:drill /tmp/drill/ -R
+
+echo "INFO:  configuring DFS storage inside Apache Drill container."
+curl -X POST \
+	-H "Content-Type: application/json" \
+	-d "{\"name\":\"dfs\",\"config\":{\"type\":\"file\",\"enabled\":true,\"connection\":\"file:///\",\"workspaces\":{\"root\":{\"location\":\"/\",\"writable\":false,\"defaultInputFormat\":null},\"tmp\":{\"location\":\"/tmp\",\"writable\":true,\"defaultIn putFormat\":null},\"views\":{\"location\":\"/data\",\"writable\":true,\"defaultInputFormat\":null},\"parquet\":{\"location\":\"/mnt/backup\",\"writable\":true,\"defaultInputFormat\":null}},\"formats\":{\"psv\":{\"type\":\"text\",\"extensions\":[\"tbl\"],\"delimiter\":\"|\"},\"csv\":{\"type\":\"text\",\"extensions\":[\"csv\"],\"delimiter\":\",\"},\"tsv\":{\"type\":\"text\",\"extensions\":[\"tsv\"],\"delimiter\":\"\\t\"},\"parquet\":{\"type\":\"parquet\"},\"json\":{\"type\":\"json\",\"extensions\":[\"json\"]},\"avro\":{\"type\":\"avro\"},\"sequencefile\":{\"type\":\"sequencefile\",\"extensions\":[\"seq\"]},\"csvh\":{\"type\":\"text\",\"extensions\":[\"csvh\"],\"extractHeader\":\"true\",\"delimiter\":\",\"}}}}" \
+	${DRILL_CONFIG_URL}/storage/dfs.json
+
+echo "INFO: configuring Kudu storage inside Apache Drill container."
+curl -X POST \
+	-H "Content-Type: application/json" \
+	-d "{\"name\":\"kudu\", \"config\": {\"type\": \"kudu\", \"masterAddresses\": \"${KUDU_MASTER_INT_IP}\", \"operationTimeoutMs\": \"600000\", \"optimizerMaxNonPrimaryKeyAlternatives\": \"20\", \"allUnsignedINT8\": true, \"allUnsignedINT16\": true, \"enabled\": \"true\"}}" \
+	${DRILL_CONFIG_URL}/storage/kudu.json
+
+echo "INFO: configuring Netdisco storage inside Apache Drill container."
+curl -X POST \
+	-H "Content-Type: application/json" \
+	-d "{\"name\":\"netdisco\", \"config\": {\"type\": \"jdbc\", \"url\": \"jdbc:postgresql://${NETDISCO_DB_HOST}:${NETDISCO_DB_PORT}/${NETDISCO_DB_NAME}\", \"driver\": \"org.postgresql.Driver\", \"username\": \"${NETDISCO_DB_USER}\", \"password\": \"${NETDISCO_DB_PASS}\", \"enabled\": \"true\"}}" \
+	${DRILL_CONFIG_URL}/storage/netdisco.json
+
+echo "INFO: enabling multithreading in Drill queries."
+curl -X POST \
+	-H "Content-Type: application/json" \
+	-d "{\"queryType\":\"SQL\",\"query\":\"ALTER SYSTEM SET planner.slice_target = 1000\"}" \
+	${DRILL_CONFIG_URL}/query.json
+
+echo "INFO: enabling fallback for hash aggregations."
+curl -X POST \
+	-H "Content-Type: application/json" \
+	-d "{\"queryType\":\"SQL\",\"query\":\"ALERT SYSTEM SET \`drill.exec.hashagg.fallback.enabled\` = true\"}" \
+	${DRILL_CONFIG_URL}/query.json
+
+echo "INFO: Drill container started."

dockerization/files/wait-for.sh

@@ -0,0 +1,28 @@
+#!/bin/bash
+
+HOST=$1
+PORT=$2
+STARTUP_TIMEOUT=$3
+
+TIMEOUT=${STARTUP_TIMEOUT:-30}
+SLEEP_TIME=1
+
+printf "Waiting for ${HOST}:${PORT} to be accessible..."
+
+counter=0
+result="1"
+while [ "$result" != 200 ]; do
+    result=$(curl -s -o /dev/null -w "%{http_code}" ${DRILL_CONFIG_URL}/storage/dfs.json)
+
+    counter=$[$counter + $SLEEP_TIME]
+    if [ $counter -gt $TIMEOUT ]; then
+           printf " FAIL\n"
+           echo "The service ${HOST}:${PORT} is not accessible after ${TIMEOUT} sec timeout. Exiting."
+           exit 1
+    fi
+    [[ "$result" != 200 ]] && sleep $SLEEP_TIME;
+    printf "%s" "." ;
+done
+printf " OK\n"
+
+exit 0