forked from D-Net/dnet-hadoop
Merge pull request 'Miscellaneous updates to the copying operation to Impala Cluster' (#440) from antonis.lempesis/dnet-hadoop:beta into beta
Reviewed-on: D-Net/dnet-hadoop#440
This commit is contained in:
commit
62716141c5
|
@ -72,6 +72,8 @@ function copydb() {
|
||||||
rm -f error.log
|
rm -f error.log
|
||||||
if [[ SHOULD_EXIT_WHOLE_SCRIPT_UPON_ERROR -eq 1 ]]; then
|
if [[ SHOULD_EXIT_WHOLE_SCRIPT_UPON_ERROR -eq 1 ]]; then
|
||||||
exit 2
|
exit 2
|
||||||
|
else
|
||||||
|
return 2
|
||||||
fi
|
fi
|
||||||
fi
|
fi
|
||||||
|
|
||||||
|
@ -90,19 +92,30 @@ function copydb() {
|
||||||
-pb \
|
-pb \
|
||||||
${OCEAN_HDFS_NODE}/user/hive/warehouse/${db}.db ${IMPALA_HDFS_DB_BASE_PATH}
|
${OCEAN_HDFS_NODE}/user/hive/warehouse/${db}.db ${IMPALA_HDFS_DB_BASE_PATH}
|
||||||
|
|
||||||
# Check the exit status of the "hadoop distcp" command.
|
if [ $? -eq 0 ]; then # Check the exit status of the "hadoop distcp" command.
|
||||||
if [ $? -eq 0 ]; then
|
echo -e "\nSuccessfully copied the files of '${db}' from Ocean to Impala cluster.\n"
|
||||||
echo -e "\nSuccessfully copied the files of '${db}'.\n"
|
|
||||||
else
|
else
|
||||||
echo -e "\n\nERROR: FAILED TO TRANSFER THE FILES OF '${db}', WITH 'hadoop distcp'. GOT EXIT STATUS: $?\n\n"
|
echo -e "\n\nERROR: FAILED TO TRANSFER THE FILES OF '${db}', WITH 'hadoop distcp'. GOT EXIT STATUS: $?\n\n"
|
||||||
rm -f error.log
|
rm -f error.log
|
||||||
if [[ SHOULD_EXIT_WHOLE_SCRIPT_UPON_ERROR -eq 1 ]]; then
|
if [[ SHOULD_EXIT_WHOLE_SCRIPT_UPON_ERROR -eq 1 ]]; then
|
||||||
exit 3
|
exit 3
|
||||||
|
else
|
||||||
|
return 3
|
||||||
fi
|
fi
|
||||||
fi
|
fi
|
||||||
|
|
||||||
# In case we ever use this script for a writable DB (using inserts/updates), we should perform the following costly operation as well..
|
# Give WRITE and EXECUTE permissions to the DBs' directory only, in order to be able to create more tables later, on top of that DB.
|
||||||
#hdfs dfs -conf ${IMPALA_CONFIG_FILE} -chmod -R 777 ${TEMP_SUBDIR_FULLPATH}/${db}.db
|
hdfs dfs -conf ${IMPALA_CONFIG_FILE} -chmod u+wx ${IMPALA_HDFS_DB_BASE_PATH}/${db}.db
|
||||||
|
# In case we ever use this script for a writable DB (using inserts/updates), we should perform the costly recursive operation as well, using the "-R" param.
|
||||||
|
if [ $? -ne 0 ]; then # Check the exit status..
|
||||||
|
echo -e "\n\nERROR: FAILED TO ASSIGN WRITE AND EXECUTE PERMISSIONS TO THE DIRECTORY OF DB: '${db}'. GOT EXIT STATUS: $?\n\n"
|
||||||
|
rm -f error.log
|
||||||
|
if [[ SHOULD_EXIT_WHOLE_SCRIPT_UPON_ERROR -eq 1 ]]; then
|
||||||
|
exit 4
|
||||||
|
else
|
||||||
|
return 4
|
||||||
|
fi
|
||||||
|
fi
|
||||||
|
|
||||||
echo -e "\nCreating schema for db: '${db}'\n"
|
echo -e "\nCreating schema for db: '${db}'\n"
|
||||||
|
|
||||||
|
@ -131,7 +144,7 @@ function copydb() {
|
||||||
if [ -z "$CURRENT_PRQ_FILE" ]; then # If there is not parquet-file inside.
|
if [ -z "$CURRENT_PRQ_FILE" ]; then # If there is not parquet-file inside.
|
||||||
echo -e "\nERROR: THE TABLE \"${i}\" HAD NO FILES TO GET THE SCHEMA FROM! IT'S EMPTY!\n\n"
|
echo -e "\nERROR: THE TABLE \"${i}\" HAD NO FILES TO GET THE SCHEMA FROM! IT'S EMPTY!\n\n"
|
||||||
if [[ SHOULD_EXIT_WHOLE_SCRIPT_UPON_ERROR -eq 1 ]]; then
|
if [[ SHOULD_EXIT_WHOLE_SCRIPT_UPON_ERROR -eq 1 ]]; then
|
||||||
exit 4
|
exit 5
|
||||||
fi
|
fi
|
||||||
else
|
else
|
||||||
impala-shell --user ${HADOOP_USER_NAME} -i ${IMPALA_HOSTNAME} -q "create table ${db}.${i} like parquet '${CURRENT_PRQ_FILE}' stored as parquet;" |& tee error.log
|
impala-shell --user ${HADOOP_USER_NAME} -i ${IMPALA_HOSTNAME} -q "create table ${db}.${i} like parquet '${CURRENT_PRQ_FILE}' stored as parquet;" |& tee error.log
|
||||||
|
@ -139,7 +152,7 @@ function copydb() {
|
||||||
if [ -n "$log_errors" ]; then
|
if [ -n "$log_errors" ]; then
|
||||||
echo -e "\n\nERROR: THERE WAS A PROBLEM WHEN CREATING TABLE '${i}'!\n\n"
|
echo -e "\n\nERROR: THERE WAS A PROBLEM WHEN CREATING TABLE '${i}'!\n\n"
|
||||||
if [[ SHOULD_EXIT_WHOLE_SCRIPT_UPON_ERROR -eq 1 ]]; then
|
if [[ SHOULD_EXIT_WHOLE_SCRIPT_UPON_ERROR -eq 1 ]]; then
|
||||||
exit 5
|
exit 6
|
||||||
fi
|
fi
|
||||||
fi
|
fi
|
||||||
fi
|
fi
|
||||||
|
@ -185,7 +198,7 @@ function copydb() {
|
||||||
if [[ $new_num_of_views_to_retry -eq $previous_num_of_views_to_retry ]]; then
|
if [[ $new_num_of_views_to_retry -eq $previous_num_of_views_to_retry ]]; then
|
||||||
echo -e "\n\nERROR: THE NUMBER OF VIEWS TO RETRY HAS NOT BEEN REDUCED! THE SCRIPT IS LIKELY GOING TO AN INFINITE-LOOP! EXITING..\n\n"
|
echo -e "\n\nERROR: THE NUMBER OF VIEWS TO RETRY HAS NOT BEEN REDUCED! THE SCRIPT IS LIKELY GOING TO AN INFINITE-LOOP! EXITING..\n\n"
|
||||||
if [[ SHOULD_EXIT_WHOLE_SCRIPT_UPON_ERROR -eq 1 ]]; then
|
if [[ SHOULD_EXIT_WHOLE_SCRIPT_UPON_ERROR -eq 1 ]]; then
|
||||||
exit 6
|
exit 7
|
||||||
fi
|
fi
|
||||||
elif [[ $new_num_of_views_to_retry -gt 0 ]]; then
|
elif [[ $new_num_of_views_to_retry -gt 0 ]]; then
|
||||||
echo -e "\nTo be retried \"create_view_statements\" (${new_num_of_views_to_retry}):\n\n${all_create_view_statements[@]}\n"
|
echo -e "\nTo be retried \"create_view_statements\" (${new_num_of_views_to_retry}):\n\n${all_create_view_statements[@]}\n"
|
||||||
|
@ -215,7 +228,7 @@ function copydb() {
|
||||||
echo -e "\n\nERROR: 1 OR MORE ENTITIES OF DB '${db}' FAILED TO BE COPIED TO IMPALA CLUSTER!\n\n"
|
echo -e "\n\nERROR: 1 OR MORE ENTITIES OF DB '${db}' FAILED TO BE COPIED TO IMPALA CLUSTER!\n\n"
|
||||||
rm -f error.log
|
rm -f error.log
|
||||||
if [[ SHOULD_EXIT_WHOLE_SCRIPT_UPON_ERROR -eq 1 ]]; then
|
if [[ SHOULD_EXIT_WHOLE_SCRIPT_UPON_ERROR -eq 1 ]]; then
|
||||||
exit 7
|
exit 8
|
||||||
fi
|
fi
|
||||||
fi
|
fi
|
||||||
|
|
||||||
|
|
|
@ -72,6 +72,8 @@ function copydb() {
|
||||||
rm -f error.log
|
rm -f error.log
|
||||||
if [[ SHOULD_EXIT_WHOLE_SCRIPT_UPON_ERROR -eq 1 ]]; then
|
if [[ SHOULD_EXIT_WHOLE_SCRIPT_UPON_ERROR -eq 1 ]]; then
|
||||||
exit 2
|
exit 2
|
||||||
|
else
|
||||||
|
return 2
|
||||||
fi
|
fi
|
||||||
fi
|
fi
|
||||||
|
|
||||||
|
@ -90,19 +92,30 @@ function copydb() {
|
||||||
-pb \
|
-pb \
|
||||||
${OCEAN_HDFS_NODE}/user/hive/warehouse/${db}.db ${IMPALA_HDFS_DB_BASE_PATH}
|
${OCEAN_HDFS_NODE}/user/hive/warehouse/${db}.db ${IMPALA_HDFS_DB_BASE_PATH}
|
||||||
|
|
||||||
# Check the exit status of the "hadoop distcp" command.
|
if [ $? -eq 0 ]; then # Check the exit status of the "hadoop distcp" command.
|
||||||
if [ $? -eq 0 ]; then
|
echo -e "\nSuccessfully copied the files of '${db}' from Ocean to Impala cluster.\n"
|
||||||
echo -e "\nSuccessfully copied the files of '${db}'.\n"
|
|
||||||
else
|
else
|
||||||
echo -e "\n\nERROR: FAILED TO TRANSFER THE FILES OF '${db}', WITH 'hadoop distcp'. GOT EXIT STATUS: $?\n\n"
|
echo -e "\n\nERROR: FAILED TO TRANSFER THE FILES OF '${db}', WITH 'hadoop distcp'. GOT EXIT STATUS: $?\n\n"
|
||||||
rm -f error.log
|
rm -f error.log
|
||||||
if [[ SHOULD_EXIT_WHOLE_SCRIPT_UPON_ERROR -eq 1 ]]; then
|
if [[ SHOULD_EXIT_WHOLE_SCRIPT_UPON_ERROR -eq 1 ]]; then
|
||||||
exit 3
|
exit 3
|
||||||
|
else
|
||||||
|
return 3
|
||||||
fi
|
fi
|
||||||
fi
|
fi
|
||||||
|
|
||||||
# In case we ever use this script for a writable DB (using inserts/updates), we should perform the following costly operation as well..
|
# Give WRITE and EXECUTE permissions to the DBs' directory only, in order to be able to create more tables later, on top of that DB.
|
||||||
#hdfs dfs -conf ${IMPALA_CONFIG_FILE} -chmod -R 777 ${TEMP_SUBDIR_FULLPATH}/${db}.db
|
hdfs dfs -conf ${IMPALA_CONFIG_FILE} -chmod u+wx ${IMPALA_HDFS_DB_BASE_PATH}/${db}.db
|
||||||
|
# In case we ever use this script for a writable DB (using inserts/updates), we should perform the costly recursive operation as well, using the "-R" param.
|
||||||
|
if [ $? -ne 0 ]; then # Check the exit status..
|
||||||
|
echo -e "\n\nERROR: FAILED TO ASSIGN WRITE AND EXECUTE PERMISSIONS TO THE DIRECTORY OF DB: '${db}'. GOT EXIT STATUS: $?\n\n"
|
||||||
|
rm -f error.log
|
||||||
|
if [[ SHOULD_EXIT_WHOLE_SCRIPT_UPON_ERROR -eq 1 ]]; then
|
||||||
|
exit 4
|
||||||
|
else
|
||||||
|
return 4
|
||||||
|
fi
|
||||||
|
fi
|
||||||
|
|
||||||
echo -e "\nCreating schema for db: '${db}'\n"
|
echo -e "\nCreating schema for db: '${db}'\n"
|
||||||
|
|
||||||
|
@ -131,7 +144,7 @@ function copydb() {
|
||||||
if [ -z "$CURRENT_PRQ_FILE" ]; then # If there is not parquet-file inside.
|
if [ -z "$CURRENT_PRQ_FILE" ]; then # If there is not parquet-file inside.
|
||||||
echo -e "\nERROR: THE TABLE \"${i}\" HAD NO FILES TO GET THE SCHEMA FROM! IT'S EMPTY!\n\n"
|
echo -e "\nERROR: THE TABLE \"${i}\" HAD NO FILES TO GET THE SCHEMA FROM! IT'S EMPTY!\n\n"
|
||||||
if [[ SHOULD_EXIT_WHOLE_SCRIPT_UPON_ERROR -eq 1 ]]; then
|
if [[ SHOULD_EXIT_WHOLE_SCRIPT_UPON_ERROR -eq 1 ]]; then
|
||||||
exit 4
|
exit 5
|
||||||
fi
|
fi
|
||||||
else
|
else
|
||||||
impala-shell --user ${HADOOP_USER_NAME} -i ${IMPALA_HOSTNAME} -q "create table ${db}.${i} like parquet '${CURRENT_PRQ_FILE}' stored as parquet;" |& tee error.log
|
impala-shell --user ${HADOOP_USER_NAME} -i ${IMPALA_HOSTNAME} -q "create table ${db}.${i} like parquet '${CURRENT_PRQ_FILE}' stored as parquet;" |& tee error.log
|
||||||
|
@ -139,7 +152,7 @@ function copydb() {
|
||||||
if [ -n "$log_errors" ]; then
|
if [ -n "$log_errors" ]; then
|
||||||
echo -e "\n\nERROR: THERE WAS A PROBLEM WHEN CREATING TABLE '${i}'!\n\n"
|
echo -e "\n\nERROR: THERE WAS A PROBLEM WHEN CREATING TABLE '${i}'!\n\n"
|
||||||
if [[ SHOULD_EXIT_WHOLE_SCRIPT_UPON_ERROR -eq 1 ]]; then
|
if [[ SHOULD_EXIT_WHOLE_SCRIPT_UPON_ERROR -eq 1 ]]; then
|
||||||
exit 5
|
exit 6
|
||||||
fi
|
fi
|
||||||
fi
|
fi
|
||||||
fi
|
fi
|
||||||
|
@ -185,7 +198,7 @@ function copydb() {
|
||||||
if [[ $new_num_of_views_to_retry -eq $previous_num_of_views_to_retry ]]; then
|
if [[ $new_num_of_views_to_retry -eq $previous_num_of_views_to_retry ]]; then
|
||||||
echo -e "\n\nERROR: THE NUMBER OF VIEWS TO RETRY HAS NOT BEEN REDUCED! THE SCRIPT IS LIKELY GOING TO AN INFINITE-LOOP! EXITING..\n\n"
|
echo -e "\n\nERROR: THE NUMBER OF VIEWS TO RETRY HAS NOT BEEN REDUCED! THE SCRIPT IS LIKELY GOING TO AN INFINITE-LOOP! EXITING..\n\n"
|
||||||
if [[ SHOULD_EXIT_WHOLE_SCRIPT_UPON_ERROR -eq 1 ]]; then
|
if [[ SHOULD_EXIT_WHOLE_SCRIPT_UPON_ERROR -eq 1 ]]; then
|
||||||
exit 6
|
exit 7
|
||||||
fi
|
fi
|
||||||
elif [[ $new_num_of_views_to_retry -gt 0 ]]; then
|
elif [[ $new_num_of_views_to_retry -gt 0 ]]; then
|
||||||
echo -e "\nTo be retried \"create_view_statements\" (${new_num_of_views_to_retry}):\n\n${all_create_view_statements[@]}\n"
|
echo -e "\nTo be retried \"create_view_statements\" (${new_num_of_views_to_retry}):\n\n${all_create_view_statements[@]}\n"
|
||||||
|
@ -215,7 +228,7 @@ function copydb() {
|
||||||
echo -e "\n\nERROR: 1 OR MORE ENTITIES OF DB '${db}' FAILED TO BE COPIED TO IMPALA CLUSTER!\n\n"
|
echo -e "\n\nERROR: 1 OR MORE ENTITIES OF DB '${db}' FAILED TO BE COPIED TO IMPALA CLUSTER!\n\n"
|
||||||
rm -f error.log
|
rm -f error.log
|
||||||
if [[ SHOULD_EXIT_WHOLE_SCRIPT_UPON_ERROR -eq 1 ]]; then
|
if [[ SHOULD_EXIT_WHOLE_SCRIPT_UPON_ERROR -eq 1 ]]; then
|
||||||
exit 7
|
exit 8
|
||||||
fi
|
fi
|
||||||
fi
|
fi
|
||||||
|
|
||||||
|
|
|
@ -72,6 +72,8 @@ function copydb() {
|
||||||
rm -f error.log
|
rm -f error.log
|
||||||
if [[ SHOULD_EXIT_WHOLE_SCRIPT_UPON_ERROR -eq 1 ]]; then
|
if [[ SHOULD_EXIT_WHOLE_SCRIPT_UPON_ERROR -eq 1 ]]; then
|
||||||
exit 2
|
exit 2
|
||||||
|
else
|
||||||
|
return 2
|
||||||
fi
|
fi
|
||||||
fi
|
fi
|
||||||
|
|
||||||
|
@ -90,19 +92,30 @@ function copydb() {
|
||||||
-pb \
|
-pb \
|
||||||
${OCEAN_HDFS_NODE}/user/hive/warehouse/${db}.db ${IMPALA_HDFS_DB_BASE_PATH}
|
${OCEAN_HDFS_NODE}/user/hive/warehouse/${db}.db ${IMPALA_HDFS_DB_BASE_PATH}
|
||||||
|
|
||||||
# Check the exit status of the "hadoop distcp" command.
|
if [ $? -eq 0 ]; then # Check the exit status of the "hadoop distcp" command.
|
||||||
if [ $? -eq 0 ]; then
|
echo -e "\nSuccessfully copied the files of '${db}' from Ocean to Impala cluster.\n"
|
||||||
echo -e "\nSuccessfully copied the files of '${db}'.\n"
|
|
||||||
else
|
else
|
||||||
echo -e "\n\nERROR: FAILED TO TRANSFER THE FILES OF '${db}', WITH 'hadoop distcp'. GOT EXIT STATUS: $?\n\n"
|
echo -e "\n\nERROR: FAILED TO TRANSFER THE FILES OF '${db}', WITH 'hadoop distcp'. GOT EXIT STATUS: $?\n\n"
|
||||||
rm -f error.log
|
rm -f error.log
|
||||||
if [[ SHOULD_EXIT_WHOLE_SCRIPT_UPON_ERROR -eq 1 ]]; then
|
if [[ SHOULD_EXIT_WHOLE_SCRIPT_UPON_ERROR -eq 1 ]]; then
|
||||||
exit 3
|
exit 3
|
||||||
|
else
|
||||||
|
return 3
|
||||||
fi
|
fi
|
||||||
fi
|
fi
|
||||||
|
|
||||||
# In case we ever use this script for a writable DB (using inserts/updates), we should perform the following costly operation as well..
|
# Give WRITE and EXECUTE permissions to the DBs' directory only, in order to be able to create more tables later, on top of that DB.
|
||||||
#hdfs dfs -conf ${IMPALA_CONFIG_FILE} -chmod -R 777 ${TEMP_SUBDIR_FULLPATH}/${db}.db
|
hdfs dfs -conf ${IMPALA_CONFIG_FILE} -chmod u+wx ${IMPALA_HDFS_DB_BASE_PATH}/${db}.db
|
||||||
|
# In case we ever use this script for a writable DB (using inserts/updates), we should perform the costly recursive operation as well, using the "-R" param.
|
||||||
|
if [ $? -ne 0 ]; then # Check the exit status..
|
||||||
|
echo -e "\n\nERROR: FAILED TO ASSIGN WRITE AND EXECUTE PERMISSIONS TO THE DIRECTORY OF DB: '${db}'. GOT EXIT STATUS: $?\n\n"
|
||||||
|
rm -f error.log
|
||||||
|
if [[ SHOULD_EXIT_WHOLE_SCRIPT_UPON_ERROR -eq 1 ]]; then
|
||||||
|
exit 4
|
||||||
|
else
|
||||||
|
return 4
|
||||||
|
fi
|
||||||
|
fi
|
||||||
|
|
||||||
echo -e "\nCreating schema for db: '${db}'\n"
|
echo -e "\nCreating schema for db: '${db}'\n"
|
||||||
|
|
||||||
|
@ -131,7 +144,7 @@ function copydb() {
|
||||||
if [ -z "$CURRENT_PRQ_FILE" ]; then # If there is not parquet-file inside.
|
if [ -z "$CURRENT_PRQ_FILE" ]; then # If there is not parquet-file inside.
|
||||||
echo -e "\nERROR: THE TABLE \"${i}\" HAD NO FILES TO GET THE SCHEMA FROM! IT'S EMPTY!\n\n"
|
echo -e "\nERROR: THE TABLE \"${i}\" HAD NO FILES TO GET THE SCHEMA FROM! IT'S EMPTY!\n\n"
|
||||||
if [[ SHOULD_EXIT_WHOLE_SCRIPT_UPON_ERROR -eq 1 ]]; then
|
if [[ SHOULD_EXIT_WHOLE_SCRIPT_UPON_ERROR -eq 1 ]]; then
|
||||||
exit 4
|
exit 5
|
||||||
fi
|
fi
|
||||||
else
|
else
|
||||||
impala-shell --user ${HADOOP_USER_NAME} -i ${IMPALA_HOSTNAME} -q "create table ${db}.${i} like parquet '${CURRENT_PRQ_FILE}' stored as parquet;" |& tee error.log
|
impala-shell --user ${HADOOP_USER_NAME} -i ${IMPALA_HOSTNAME} -q "create table ${db}.${i} like parquet '${CURRENT_PRQ_FILE}' stored as parquet;" |& tee error.log
|
||||||
|
@ -139,7 +152,7 @@ function copydb() {
|
||||||
if [ -n "$log_errors" ]; then
|
if [ -n "$log_errors" ]; then
|
||||||
echo -e "\n\nERROR: THERE WAS A PROBLEM WHEN CREATING TABLE '${i}'!\n\n"
|
echo -e "\n\nERROR: THERE WAS A PROBLEM WHEN CREATING TABLE '${i}'!\n\n"
|
||||||
if [[ SHOULD_EXIT_WHOLE_SCRIPT_UPON_ERROR -eq 1 ]]; then
|
if [[ SHOULD_EXIT_WHOLE_SCRIPT_UPON_ERROR -eq 1 ]]; then
|
||||||
exit 5
|
exit 6
|
||||||
fi
|
fi
|
||||||
fi
|
fi
|
||||||
fi
|
fi
|
||||||
|
@ -185,7 +198,7 @@ function copydb() {
|
||||||
if [[ $new_num_of_views_to_retry -eq $previous_num_of_views_to_retry ]]; then
|
if [[ $new_num_of_views_to_retry -eq $previous_num_of_views_to_retry ]]; then
|
||||||
echo -e "\n\nERROR: THE NUMBER OF VIEWS TO RETRY HAS NOT BEEN REDUCED! THE SCRIPT IS LIKELY GOING TO AN INFINITE-LOOP! EXITING..\n\n"
|
echo -e "\n\nERROR: THE NUMBER OF VIEWS TO RETRY HAS NOT BEEN REDUCED! THE SCRIPT IS LIKELY GOING TO AN INFINITE-LOOP! EXITING..\n\n"
|
||||||
if [[ SHOULD_EXIT_WHOLE_SCRIPT_UPON_ERROR -eq 1 ]]; then
|
if [[ SHOULD_EXIT_WHOLE_SCRIPT_UPON_ERROR -eq 1 ]]; then
|
||||||
exit 6
|
exit 7
|
||||||
fi
|
fi
|
||||||
elif [[ $new_num_of_views_to_retry -gt 0 ]]; then
|
elif [[ $new_num_of_views_to_retry -gt 0 ]]; then
|
||||||
echo -e "\nTo be retried \"create_view_statements\" (${new_num_of_views_to_retry}):\n\n${all_create_view_statements[@]}\n"
|
echo -e "\nTo be retried \"create_view_statements\" (${new_num_of_views_to_retry}):\n\n${all_create_view_statements[@]}\n"
|
||||||
|
@ -215,7 +228,7 @@ function copydb() {
|
||||||
echo -e "\n\nERROR: 1 OR MORE ENTITIES OF DB '${db}' FAILED TO BE COPIED TO IMPALA CLUSTER!\n\n"
|
echo -e "\n\nERROR: 1 OR MORE ENTITIES OF DB '${db}' FAILED TO BE COPIED TO IMPALA CLUSTER!\n\n"
|
||||||
rm -f error.log
|
rm -f error.log
|
||||||
if [[ SHOULD_EXIT_WHOLE_SCRIPT_UPON_ERROR -eq 1 ]]; then
|
if [[ SHOULD_EXIT_WHOLE_SCRIPT_UPON_ERROR -eq 1 ]]; then
|
||||||
exit 7
|
exit 8
|
||||||
fi
|
fi
|
||||||
fi
|
fi
|
||||||
|
|
||||||
|
|
|
@ -74,6 +74,8 @@ function copydb() {
|
||||||
rm -f error.log
|
rm -f error.log
|
||||||
if [[ SHOULD_EXIT_WHOLE_SCRIPT_UPON_ERROR -eq 1 ]]; then
|
if [[ SHOULD_EXIT_WHOLE_SCRIPT_UPON_ERROR -eq 1 ]]; then
|
||||||
exit 2
|
exit 2
|
||||||
|
else
|
||||||
|
return 2
|
||||||
fi
|
fi
|
||||||
fi
|
fi
|
||||||
|
|
||||||
|
@ -92,19 +94,30 @@ function copydb() {
|
||||||
-pb \
|
-pb \
|
||||||
${OCEAN_HDFS_NODE}/user/hive/warehouse/${db}.db ${IMPALA_HDFS_DB_BASE_PATH}
|
${OCEAN_HDFS_NODE}/user/hive/warehouse/${db}.db ${IMPALA_HDFS_DB_BASE_PATH}
|
||||||
|
|
||||||
# Check the exit status of the "hadoop distcp" command.
|
if [ $? -eq 0 ]; then # Check the exit status of the "hadoop distcp" command.
|
||||||
if [ $? -eq 0 ]; then
|
echo -e "\nSuccessfully copied the files of '${db}' from Ocean to Impala cluster.\n"
|
||||||
echo -e "\nSuccessfully copied the files of '${db}'.\n"
|
|
||||||
else
|
else
|
||||||
echo -e "\n\nERROR: FAILED TO TRANSFER THE FILES OF '${db}', WITH 'hadoop distcp'. GOT EXIT STATUS: $?\n\n"
|
echo -e "\n\nERROR: FAILED TO TRANSFER THE FILES OF '${db}', WITH 'hadoop distcp'. GOT EXIT STATUS: $?\n\n"
|
||||||
rm -f error.log
|
rm -f error.log
|
||||||
if [[ SHOULD_EXIT_WHOLE_SCRIPT_UPON_ERROR -eq 1 ]]; then
|
if [[ SHOULD_EXIT_WHOLE_SCRIPT_UPON_ERROR -eq 1 ]]; then
|
||||||
exit 3
|
exit 3
|
||||||
|
else
|
||||||
|
return 3
|
||||||
fi
|
fi
|
||||||
fi
|
fi
|
||||||
|
|
||||||
# In case we ever use this script for a writable DB (using inserts/updates), we should perform the following costly operation as well..
|
# Give WRITE and EXECUTE permissions to the DBs' directory only, in order to be able to create more tables later, on top of that DB.
|
||||||
#hdfs dfs -conf ${IMPALA_CONFIG_FILE} -chmod -R 777 ${TEMP_SUBDIR_FULLPATH}/${db}.db
|
hdfs dfs -conf ${IMPALA_CONFIG_FILE} -chmod u+wx ${IMPALA_HDFS_DB_BASE_PATH}/${db}.db
|
||||||
|
# In case we ever use this script for a writable DB (using inserts/updates), we should perform the costly recursive operation as well, using the "-R" param.
|
||||||
|
if [ $? -ne 0 ]; then # Check the exit status..
|
||||||
|
echo -e "\n\nERROR: FAILED TO ASSIGN WRITE AND EXECUTE PERMISSIONS TO THE DIRECTORY OF DB: '${db}'. GOT EXIT STATUS: $?\n\n"
|
||||||
|
rm -f error.log
|
||||||
|
if [[ SHOULD_EXIT_WHOLE_SCRIPT_UPON_ERROR -eq 1 ]]; then
|
||||||
|
exit 4
|
||||||
|
else
|
||||||
|
return 4
|
||||||
|
fi
|
||||||
|
fi
|
||||||
|
|
||||||
echo -e "\nCreating schema for db: '${db}'\n"
|
echo -e "\nCreating schema for db: '${db}'\n"
|
||||||
|
|
||||||
|
@ -133,7 +146,7 @@ function copydb() {
|
||||||
if [ -z "$CURRENT_PRQ_FILE" ]; then # If there is not parquet-file inside.
|
if [ -z "$CURRENT_PRQ_FILE" ]; then # If there is not parquet-file inside.
|
||||||
echo -e "\nERROR: THE TABLE \"${i}\" HAD NO FILES TO GET THE SCHEMA FROM! IT'S EMPTY!\n\n"
|
echo -e "\nERROR: THE TABLE \"${i}\" HAD NO FILES TO GET THE SCHEMA FROM! IT'S EMPTY!\n\n"
|
||||||
if [[ SHOULD_EXIT_WHOLE_SCRIPT_UPON_ERROR -eq 1 ]]; then
|
if [[ SHOULD_EXIT_WHOLE_SCRIPT_UPON_ERROR -eq 1 ]]; then
|
||||||
exit 4
|
exit 5
|
||||||
fi
|
fi
|
||||||
else
|
else
|
||||||
impala-shell --user ${HADOOP_USER_NAME} -i ${IMPALA_HOSTNAME} -q "create table ${db}.${i} like parquet '${CURRENT_PRQ_FILE}' stored as parquet;" |& tee error.log
|
impala-shell --user ${HADOOP_USER_NAME} -i ${IMPALA_HOSTNAME} -q "create table ${db}.${i} like parquet '${CURRENT_PRQ_FILE}' stored as parquet;" |& tee error.log
|
||||||
|
@ -141,7 +154,7 @@ function copydb() {
|
||||||
if [ -n "$log_errors" ]; then
|
if [ -n "$log_errors" ]; then
|
||||||
echo -e "\n\nERROR: THERE WAS A PROBLEM WHEN CREATING TABLE '${i}'!\n\n"
|
echo -e "\n\nERROR: THERE WAS A PROBLEM WHEN CREATING TABLE '${i}'!\n\n"
|
||||||
if [[ SHOULD_EXIT_WHOLE_SCRIPT_UPON_ERROR -eq 1 ]]; then
|
if [[ SHOULD_EXIT_WHOLE_SCRIPT_UPON_ERROR -eq 1 ]]; then
|
||||||
exit 5
|
exit 6
|
||||||
fi
|
fi
|
||||||
fi
|
fi
|
||||||
fi
|
fi
|
||||||
|
@ -187,7 +200,7 @@ function copydb() {
|
||||||
if [[ $new_num_of_views_to_retry -eq $previous_num_of_views_to_retry ]]; then
|
if [[ $new_num_of_views_to_retry -eq $previous_num_of_views_to_retry ]]; then
|
||||||
echo -e "\n\nERROR: THE NUMBER OF VIEWS TO RETRY HAS NOT BEEN REDUCED! THE SCRIPT IS LIKELY GOING TO AN INFINITE-LOOP! EXITING..\n\n"
|
echo -e "\n\nERROR: THE NUMBER OF VIEWS TO RETRY HAS NOT BEEN REDUCED! THE SCRIPT IS LIKELY GOING TO AN INFINITE-LOOP! EXITING..\n\n"
|
||||||
if [[ SHOULD_EXIT_WHOLE_SCRIPT_UPON_ERROR -eq 1 ]]; then
|
if [[ SHOULD_EXIT_WHOLE_SCRIPT_UPON_ERROR -eq 1 ]]; then
|
||||||
exit 6
|
exit 7
|
||||||
fi
|
fi
|
||||||
elif [[ $new_num_of_views_to_retry -gt 0 ]]; then
|
elif [[ $new_num_of_views_to_retry -gt 0 ]]; then
|
||||||
echo -e "\nTo be retried \"create_view_statements\" (${new_num_of_views_to_retry}):\n\n${all_create_view_statements[@]}\n"
|
echo -e "\nTo be retried \"create_view_statements\" (${new_num_of_views_to_retry}):\n\n${all_create_view_statements[@]}\n"
|
||||||
|
@ -217,7 +230,7 @@ function copydb() {
|
||||||
echo -e "\n\nERROR: 1 OR MORE ENTITIES OF DB '${db}' FAILED TO BE COPIED TO IMPALA CLUSTER!\n\n"
|
echo -e "\n\nERROR: 1 OR MORE ENTITIES OF DB '${db}' FAILED TO BE COPIED TO IMPALA CLUSTER!\n\n"
|
||||||
rm -f error.log
|
rm -f error.log
|
||||||
if [[ SHOULD_EXIT_WHOLE_SCRIPT_UPON_ERROR -eq 1 ]]; then
|
if [[ SHOULD_EXIT_WHOLE_SCRIPT_UPON_ERROR -eq 1 ]]; then
|
||||||
exit 7
|
exit 8
|
||||||
fi
|
fi
|
||||||
fi
|
fi
|
||||||
|
|
||||||
|
|
Loading…
Reference in New Issue