antonis.lempesis
/
dnet-hadoop
forked from D-Net/dnet-hadoop
1
1
Fork 1
Code Issues Pull Requests Projects Releases Wiki Activity

Merge pull request 'correctly selecting the active hdfs node for the impala cluster' (#405) from antonis.lempesis/dnet-hadoop:beta into beta 

Reviewed-on: D-Net/dnet-hadoop#405
This commit is contained in:
Claudio Atzori 2024-03-26 12:07:46 +01:00
parent ef52128c55 bc8c97182d
commit 730eaffc85
4 changed files with 53 additions and 8 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

16
dhp-workflows/dhp-stats-hist-snaps/src/main/resources/eu/dnetlib/dhp/oa/graph/stats-hist-snaps/oozie_app/copyDataToImpalaCluster.sh
View File

@ -8,12 +8,24 @@ fi
export HADOOP_USER_NAME=$2 export HADOOP_USER_NAME=$2
IMPALA_HDFS_NODE=''
if hdfs dfs -test -e hdfs://impala-cluster-mn1.openaire.eu >/dev/null 2>&1; then
IMPALA_HDFS_NODE='hdfs://impala-cluster-mn1.openaire.eu:8020'
elif hdfs dfs -test -e hdfs://impala-cluster-mn2.openaire.eu >/dev/null 2>&1; then
IMPALA_HDFS_NODE='hdfs://impala-cluster-mn2.openaire.eu:8020'
else
echo -e "\n\nPROBLEM WHEN SETTING THE HDFS-NODE FOR IMPALA CLUSTER!\n\n"
exit 1
fi
echo "Active IMPALA HDFS Node: ${IMPALA_HDFS_NODE}"
function copydb() { function copydb() {
db=$1 db=$1
FILE=("hive_wf_tmp_"$RANDOM) FILE=("hive_wf_tmp_"$RANDOM)
hdfs dfs -mkdir hdfs://impala-cluster-mn1.openaire.eu:8020/tmp/$FILE/ hdfs dfs -mkdir ${IMPALA_HDFS_NODE}/tmp/$FILE/
# change ownership to impala # change ownership to impala
# hdfs dfs -conf /etc/impala_cluster/hdfs-site.xml -chmod -R 777 /tmp/$FILE/${db}.db # hdfs dfs -conf /etc/impala_cluster/hdfs-site.xml -chmod -R 777 /tmp/$FILE/${db}.db
@ -22,7 +34,7 @@ function copydb() {
# copy the databases from ocean to impala # copy the databases from ocean to impala
echo "copying $db" echo "copying $db"
hadoop distcp -Dmapreduce.map.memory.mb=6144 -pb hdfs://nameservice1/user/hive/warehouse/${db}.db hdfs://impala-cluster-mn1.openaire.eu:8020/tmp/$FILE/ hadoop distcp -Dmapreduce.map.memory.mb=6144 -pb hdfs://nameservice1/user/hive/warehouse/${db}.db ${IMPALA_HDFS_NODE}/tmp/$FILE/
hdfs dfs -conf /etc/impala_cluster/hdfs-site.xml -chmod -R 777 /tmp/$FILE/${db}.db hdfs dfs -conf /etc/impala_cluster/hdfs-site.xml -chmod -R 777 /tmp/$FILE/${db}.db

15
dhp-workflows/dhp-stats-monitor-irish/src/main/resources/eu/dnetlib/dhp/oa/graph/stats-monitor-irish/oozie_app/copyDataToImpalaCluster.sh
View File

@ -8,11 +8,22 @@ fi
export HADOOP_USER_NAME=$2 export HADOOP_USER_NAME=$2
IMPALA_HDFS_NODE=''
if hdfs dfs -test -e hdfs://impala-cluster-mn1.openaire.eu >/dev/null 2>&1; then
IMPALA_HDFS_NODE='hdfs://impala-cluster-mn1.openaire.eu:8020'
elif hdfs dfs -test -e hdfs://impala-cluster-mn2.openaire.eu >/dev/null 2>&1; then
IMPALA_HDFS_NODE='hdfs://impala-cluster-mn2.openaire.eu:8020'
else
echo -e "\n\nPROBLEM WHEN SETTING THE HDFS-NODE FOR IMPALA CLUSTER!\n\n"
exit 1
fi
echo "Active IMPALA HDFS Node: ${IMPALA_HDFS_NODE}"
function copydb() { function copydb() {
db=$1 db=$1
FILE=("hive_wf_tmp_"$RANDOM) FILE=("hive_wf_tmp_"$RANDOM)
hdfs dfs -mkdir hdfs://impala-cluster-mn1.openaire.eu:8020/tmp/$FILE/ hdfs dfs -mkdir ${IMPALA_HDFS_NODE}/tmp/$FILE/
# change ownership to impala # change ownership to impala
# hdfs dfs -conf /etc/impala_cluster/hdfs-site.xml -chmod -R 777 /tmp/$FILE/${db}.db # hdfs dfs -conf /etc/impala_cluster/hdfs-site.xml -chmod -R 777 /tmp/$FILE/${db}.db
@ -21,7 +32,7 @@ function copydb() {
# copy the databases from ocean to impala # copy the databases from ocean to impala
echo "copying $db" echo "copying $db"
hadoop distcp -Dmapreduce.map.memory.mb=6144 -pb hdfs://nameservice1/user/hive/warehouse/${db}.db hdfs://impala-cluster-mn1.openaire.eu:8020/tmp/$FILE/ hadoop distcp -Dmapreduce.map.memory.mb=6144 -pb hdfs://nameservice1/user/hive/warehouse/${db}.db ${IMPALA_HDFS_NODE}/tmp/$FILE/
hdfs dfs -conf /etc/impala_cluster/hdfs-site.xml -chmod -R 777 /tmp/$FILE/${db}.db hdfs dfs -conf /etc/impala_cluster/hdfs-site.xml -chmod -R 777 /tmp/$FILE/${db}.db

15
dhp-workflows/dhp-stats-monitor-update/src/main/resources/eu/dnetlib/dhp/oa/graph/stats-monitor/oozie_app/copyDataToImpalaCluster.sh
View File

@ -8,6 +8,17 @@ fi
#export HADOOP_USER_NAME=$2 #export HADOOP_USER_NAME=$2
IMPALA_HDFS_NODE=''
if hdfs dfs -test -e hdfs://impala-cluster-mn1.openaire.eu >/dev/null 2>&1; then
IMPALA_HDFS_NODE='hdfs://impala-cluster-mn1.openaire.eu:8020'
elif hdfs dfs -test -e hdfs://impala-cluster-mn2.openaire.eu >/dev/null 2>&1; then
IMPALA_HDFS_NODE='hdfs://impala-cluster-mn2.openaire.eu:8020'
else
echo -e "\n\nPROBLEM WHEN SETTING THE HDFS-NODE FOR IMPALA CLUSTER!\n\n"
exit 1
fi
echo "Active IMPALA HDFS Node: ${IMPALA_HDFS_NODE}"
function copydb() { function copydb() {
export HADOOP_USER="dimitris.pierrakos" export HADOOP_USER="dimitris.pierrakos"
@ -15,7 +26,7 @@ function copydb() {
db=$1 db=$1
FILE=("hive_wf_tmp_"$RANDOM) FILE=("hive_wf_tmp_"$RANDOM)
hdfs dfs -mkdir hdfs://impala-cluster-mn1.openaire.eu:8020/tmp/$FILE/ hdfs dfs -mkdir ${IMPALA_HDFS_NODE}/tmp/$FILE/φ
# change ownership to impala # change ownership to impala
# hdfs dfs -conf /etc/impala_cluster/hdfs-site.xml -chmod -R 777 /tmp/$FILE/${db}.db # hdfs dfs -conf /etc/impala_cluster/hdfs-site.xml -chmod -R 777 /tmp/$FILE/${db}.db
@ -24,7 +35,7 @@ function copydb() {
# copy the databases from ocean to impala # copy the databases from ocean to impala
echo "copying $db" echo "copying $db"
hadoop distcp -Dmapreduce.map.memory.mb=6144 -pb hdfs://nameservice1/user/hive/warehouse/${db}.db hdfs://impala-cluster-mn1.openaire.eu:8020/tmp/$FILE/ hadoop distcp -Dmapreduce.map.memory.mb=6144 -pb hdfs://nameservice1/user/hive/warehouse/${db}.db ${IMPALA_HDFS_NODE}/tmp/$FILE/
hdfs dfs -conf /etc/impala_cluster/hdfs-site.xml -chmod -R 777 /tmp/$FILE/${db}.db hdfs dfs -conf /etc/impala_cluster/hdfs-site.xml -chmod -R 777 /tmp/$FILE/${db}.db

15
dhp-workflows/dhp-stats-update/src/main/resources/eu/dnetlib/dhp/oa/graph/stats/oozie_app/copyDataToImpalaCluster.sh
View File

@ -6,16 +6,27 @@ then
ln -sfn ${PYTHON_EGG_CACHE}${link_folder} ${link_folder} ln -sfn ${PYTHON_EGG_CACHE}${link_folder} ${link_folder}
fi fi
IMPALA_HDFS_NODE=''
if hdfs dfs -test -e hdfs://impala-cluster-mn1.openaire.eu >/dev/null 2>&1; then
IMPALA_HDFS_NODE='hdfs://impala-cluster-mn1.openaire.eu:8020'
elif hdfs dfs -test -e hdfs://impala-cluster-mn2.openaire.eu >/dev/null 2>&1; then
IMPALA_HDFS_NODE='hdfs://impala-cluster-mn2.openaire.eu:8020'
else
echo -e "\n\nPROBLEM WHEN SETTING THE HDFS-NODE FOR IMPALA CLUSTER!\n\n"
exit 1
fi
echo "Active IMPALA HDFS Node: ${IMPALA_HDFS_NODE}"
export HADOOP_USER_NAME=$6 export HADOOP_USER_NAME=$6
export PROD_USAGE_STATS_DB="openaire_prod_usage_stats" export PROD_USAGE_STATS_DB="openaire_prod_usage_stats"
function copydb() { function copydb() {
db=$1 db=$1
FILE=("hive_wf_tmp_"$RANDOM) FILE=("hive_wf_tmp_"$RANDOM)
hdfs dfs -mkdir hdfs://impala-cluster-mn1.openaire.eu:8020/tmp/$FILE/ hdfs dfs -mkdir ${IMPALA_HDFS_NODE}/tmp/$FILE/
# copy the databases from ocean to impala # copy the databases from ocean to impala
echo "copying $db" echo "copying $db"
hadoop distcp -Dmapreduce.map.memory.mb=6144 -pb hdfs://nameservice1/user/hive/warehouse/${db}.db hdfs://impala-cluster-mn1.openaire.eu:8020/tmp/$FILE/ hadoop distcp -Dmapreduce.map.memory.mb=6144 -pb hdfs://nameservice1/user/hive/warehouse/${db}.db ${IMPALA_HDFS_NODE}/tmp/$FILE/
# change ownership to impala # change ownership to impala
hdfs dfs -conf /etc/impala_cluster/hdfs-site.xml -chmod -R 777 /tmp/$FILE/${db}.db hdfs dfs -conf /etc/impala_cluster/hdfs-site.xml -chmod -R 777 /tmp/$FILE/${db}.db