Created
November 28, 2018 06:25
-
-
Save shiumachi/bdde8314f7f53270fc9cbbbf99108fab to your computer and use it in GitHub Desktop.
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
#!/bin/sh | |
source util.sh | |
TMP_DIR=/tmp | |
HADOOP_BIN_DIR=${HADOOP_HOME}/bin | |
HADOOP_COMMAND=${HADOOP_BIN_DIR}/hadoop | |
HADOOP_START_HDFS_COMMAND=`which start-dfs.sh` | |
HADOOP_STOP_HDFS_COMMAND=`which stop-dfs.sh` | |
HADOOP_NAMENODE_FORMAT_COMMAND="${HADOOP_COMMAND} namenode -format" | |
HADOOP_HDFS_COMMAND="${HADOOP_COMMAND} fs" | |
HADOOP_TMP_DIR=${TMP_DIR}/hadoop | |
# local test data path | |
DATA_DIR=${HOME}/data | |
# hdfs path | |
HDFS_TMP_DIR=/tmp | |
HDFS_TEST_DIR=/test | |
HBASE_DIR=/hbase | |
# yarn | |
YARN_LOG_HDFS_DIR=/var/log/hadoop-yarn | |
# mrv2 | |
MAPREDUCE_V2_HISTORY_DIR=/user/history | |
# basic mapreduce | |
TESTDATA_BASENAME=sherlock | |
TESTDATA_COMPRESSED=${DATA_DIR}/${TESTDATA_BASENAME}.tar.gz | |
TESTDATA=${TMP_DIR}/${TESTDATA_BASENAME} | |
HDFS_TESTDATA_DIR=${HDFS_TEST_DIR}/${TESTDATA_BASENAME} | |
# streaming script | |
STREAMING_SCRIPTS_BASENAME=streaming | |
STREAMING_SCRIPTS_COMPRESSED=${DATA_DIR}/${STREAMING_SCRIPTS_BASENAME}.tar.gz | |
STREAMING_SCRIPTS=${TMP_DIR}/${STREAMING_SCRIPTS_BASENAME} | |
HDFS_STREAMING_SCRIPTS_DIR=${HDFS_TEST_DIR}/${STREAMING_SCRIPTS_BASENAME} | |
# hive | |
TESTDATA_HIVE_BASENAME=hivedata | |
TESTDATA_HIVE_COMPRESSED=${DATA_DIR}/${TESTDATA_HIVE_BASENAME}.tar.gz | |
TESTDATA_HIVE=${TMP_DIR}/${TESTDATA_HIVE_BASENAME} | |
HDFS_TESTDATA_HIVE_DIR=${HDFS_TEST_DIR}/${TESTDATA_HIVE_BASENAME} | |
HASH=`uuidgen | cut -c 1-8` | |
usage(){ | |
printf "usage: %s -prod\n" `basename $0` 1>&2 | |
} | |
initialize(){ | |
echo "initialize()" | |
mv ${HADOOP_TMP_DIR} ${HADOOP_TMP_DIR}.${HASH} | |
} | |
initialize_dryrun(){ | |
echo "initialize()" | |
echo "mv ${HADOOP_TMP_DIR} ${HADOOP_TMP_DIR}.${HASH}" | |
} | |
namenode_format(){ | |
echo "namenode_format()" | |
${HADOOP_STOP_HDFS_COMMAND} | |
${HADOOP_NAMENODE_FORMAT_COMMAND} | |
} | |
namenode_format_dryrun(){ | |
echo "namenode_format()" | |
echo ${HADOOP_STOP_HDFS_COMMAND} | |
echo ${HADOOP_NAMENODE_FORMAT_COMMAND} | |
} | |
decompress_data(){ | |
echo "decompress_data()" | |
tar xf ${TESTDATA_COMPRESSED} -C ${TMP_DIR} | |
tar xf ${STREAMING_SCRIPTS_COMPRESSED} -C ${TMP_DIR} | |
tar xf ${TESTDATA_HIVE_COMPRESSED} -C ${TMP_DIR} | |
} | |
decompress_data_dryrun(){ | |
echo "decompress_data()" | |
echo "tar xf ${TESTDATA_COMPRESSED} -C ${TMP_DIR}" | |
echo "tar xf ${STREAMING_SCRIPTS_COMPRESSED} -C ${TMP_DIR}" | |
echo "tar xf ${TESTDATA_HIVE_COMPRESSED} -C ${TMP_DIR}" | |
} | |
start_dfs(){ | |
echo "start_dfs()" | |
${HADOOP_START_HDFS_COMMAND} | |
} | |
start_dfs_dryrun(){ | |
echo "start_dfs()" | |
echo "${HADOOP_START_HDFS_COMMAND}" | |
} | |
initialize_hdfs(){ | |
echo "initialize_hdfs()" | |
${HADOOP_HDFS_COMMAND} -mkdir ${HDFS_TMP_DIR} | |
${HADOOP_HDFS_COMMAND} -chmod 777 ${HDFS_TMP_DIR} | |
${HADOOP_HDFS_COMMAND} -mkdir ${HDFS_TEST_DIR} | |
${HADOOP_HDFS_COMMAND} -chmod 777 ${HDFS_TEST_DIR} | |
${HADOOP_HDFS_COMMAND} -mkdir ${HBASE_DIR} | |
${HADOOP_HDFS_COMMAND} -chmod 777 ${HBASE_DIR} | |
${HADOOP_HDFS_COMMAND} -mkdir ${YARN_LOG_HDFS_DIR} | |
${HADOOP_HDFS_COMMAND} -chmod 777 ${YARN_LOG_HDFS_DIR} | |
${HADOOP_HDFS_COMMAND} -mkdir ${MAPREDUCE_V2_HISTORY_DIR} | |
${HADOOP_HDFS_COMMAND} -chmod 1777 ${MAPREDUCE_V2_HISTORY_DIR} | |
} | |
initialize_hdfs_dryrun(){ | |
echo "initialize_hdfs()" | |
echo "${HADOOP_HDFS_COMMAND} -mkdir ${HDFS_TMP_DIR}" | |
echo "${HADOOP_HDFS_COMMAND} -chmod 777 ${HDFS_TMP_DIR}" | |
echo "${HADOOP_HDFS_COMMAND} -mkdir ${HDFS_TEST_DIR}" | |
echo "${HADOOP_HDFS_COMMAND} -chmod 777 ${HDFS_TEST_DIR}" | |
echo "${HADOOP_HDFS_COMMAND} -mkdir ${HBASE_DIR}" | |
echo "${HADOOP_HDFS_COMMAND} -chmod 777 ${HBASE_DIR}" | |
echo "${HADOOP_HDFS_COMMAND} -mkdir ${YARN_LOG_HDFS_DIR}" | |
echo "${HADOOP_HDFS_COMMAND} -chmod 777 ${YARN_LOG_HDFS_DIR}" | |
echo "${HADOOP_HDFS_COMMAND} -mkdir ${MAPREDUCE_V2_HISTORY_DIR}" | |
echo "${HADOOP_HDFS_COMMAND} -chmod 1777 ${MAPREDUCE_V2_HISTORY_DIR}" | |
} | |
put_testdata(){ | |
echo "put_testdata()" | |
${HADOOP_HDFS_COMMAND} -put ${TESTDATA} ${HDFS_TESTDATA_DIR} | |
${HADOOP_HDFS_COMMAND} -put ${STREAMING_SCRIPTS} ${HDFS_STREAMING_SCRIPTS_DIR} | |
${HADOOP_HDFS_COMMAND} -put ${TESTDATA_HIVE} ${HDFS_TESTDATA_HIVE_DIR} | |
} | |
put_testdata_dryrun(){ | |
echo "put_testdata()" | |
echo "${HADOOP_HDFS_COMMAND} -put ${TESTDATA} ${HDFS_TESTDATA_DIR}" | |
echo "${HADOOP_HDFS_COMMAND} -put ${STREAMING_SCRIPTS} ${HDFS_STREAMING_SCRIPTS_DIR}" | |
echo "${HADOOP_HDFS_COMMAND} -put ${TESTDATA_HIVE} ${HDFS_TESTDATA_HIVE_DIR}" | |
} | |
finalize(){ | |
echo "finalize()" | |
chmod 777 ${HADOOP_TMP_DIR} | |
} | |
finalize_dryrun(){ | |
echo "finalize()" | |
echo "chmod 777 ${HADOOP_TMP_DIR}" | |
} | |
cleanup(){ | |
echo "cleanup()" | |
mv ${TESTDATA} ${TESTDATA}.${HASH} | |
mv ${TESTDATA_HIVE} ${TESTDATA_HIVE}.${HASH} | |
} | |
cleanup_dryrun(){ | |
echo "cleanup()" | |
echo "mv ${TESTDATA} ${TESTDATA}.${HASH}" | |
echo "mv ${TESTDATA_HIVE} ${TESTDATA_HIVE}.${HASH}" | |
} | |
execute(){ | |
if [ "${DRYRUN}" == 1 ]; then | |
initialize_dryrun | |
namenode_format_dryrun | |
start_dfs_dryrun | |
initialize_hdfs_dryrun | |
decompress_data_dryrun | |
put_testdata_dryrun | |
finalize_dryrun | |
cleanup_dryrun | |
else | |
initialize | |
namenode_format | |
start_dfs | |
initialize_hdfs | |
decompress_data | |
wait_seconds 5 # wait for leaving safemode | |
put_testdata | |
finalize | |
cleanup | |
fi | |
} | |
if [ $# == 1 ]; then | |
DRYRUN=0 | |
elif [ $# == 0 ]; then | |
DRYRUN=1 | |
else | |
usage | |
exit | |
fi | |
execute |
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment