Hbase Region server summary - Useful for Troubleshooting

Shell script to collect the Hbase Region server summary report:-

It will be useful when working in command line and to monitor the Regions summary to troubleshoot the performance issues.

############Print header to Report ##############################

echo $'Region_Server\t#OnlineReg\tReq/S\tRd/S\tWr/S\tUsedHeapMB\tMemstoreMB\t#Stores\t#Store_Files\tStr_Files_size\tcompact_Prog_Pct\t#Compacting_KV\t#Curr_Compacted_KV \n' > /tmp/RS_Quick_Summary

############Get status detailed output from hbase shell##########

echo `date`

echo "Collecting RS detailed status ......"

echo "status 'detailed'" | hbase shell -n |grep -i maxHeapMB -B 1 | grep -v '\--' > /tmp/rs_status_raw_data

echo `date`

echo "Processing data....."

filename=/tmp/rs_status_raw_data

while read -r line

do

    name="$line"

############ Extract RS server name ##############################

if [[ $line == *":16020"* ]]; then

echo $line | awk '{print $1}' |cut -d':' -f 1 > /tmp/temp_rs_processing_data

rs_server=$(echo $line | awk '{print $1}' |cut -d':' -f 1)

else

############ Collect stats for each RS server from status output ############

echo $line | cut -d',' -f 1-50 --output-delimiter='' | awk -v FS=' ' -v OFS='\n' '{$1=$1}1' |grep -a -i '=' >> /tmp/temp_rs_processing_data

#

###########Metric collection######

onl_reg=$(grep -a -i numberOfOnlineRegions /tmp/temp_rs_processing_data | cut -d'=' -f 2 )

req_sec=$(grep -a -i requestsPerSecond /tmp/temp_rs_processing_data | cut -d'=' -f 2)

read_sec=$(grep -a -i readRequestsCount    /tmp/temp_rs_processing_data| cut -d'=' -f 2)

write_sec=$(grep -a -i writeRequestsCount /tmp/temp_rs_processing_data| cut -d'=' -f 2)

used_heap=$(grep -a -i usedHeapMB /tmp/temp_rs_processing_data| cut -d'=' -f 2)

memstore=$(grep -a -i memstoreSizeMB    /tmp/temp_rs_processing_data| cut -d'=' -f 2)

store_num=$(grep -a -i numberOfStores    /tmp/temp_rs_processing_data| cut -d'=' -f 2)

store_file_num=$(grep -a -i numberOfStorefiles /tmp/temp_rs_processing_data | cut -d'=' -f 2)

store_size=$(grep -a -i storefileSizeMB     /tmp/temp_rs_processing_data| cut -d'=' -f 2)

comp_pct=$(grep -a -i compactionProgressPct     /tmp/temp_rs_processing_data| cut -d'=' -f 2)

comp_in_pgrs=$(grep -a -i totalCompactingKVs /tmp/temp_rs_processing_data| cut -d'=' -f 2)

comp_complete=$(grep -a -i currentCompactedKVs /tmp/temp_rs_processing_data| cut -d'=' -f 2)

#############Print Collected information in a tab separated format############

echo $rs_server$'\t'$onl_reg$'\t'$req_sec$'\t'$read_sec$'\t'$write_sec$'\t'$used_heap$'\t'$memstore$'\t'$store_num$'\t'$store_file_num$'\t'$store_size$'\t'$comp_pct$'\t'$comp_in_pgrs$'\t'$comp_complete >> /tmp/RS_Quick_Summary

fi

done < "$filename"

############Displaying data in tab separated with formatted ##################

column -t /tmp/RS_Quick_Summary

##########Clearing Tempfiles created ####################

rm /tmp/temp_rs_processing_data /tmp/RS_Quick_Summary /tmp/rs_status_raw_data

##########################

Save it in a shell script for example  "RS_Quick_Summary.sh"  and the output will look the following.

Run this script as hbase user.  If kerberos enabled in the cluster , do a kinit before running this script.