我们以前使用过的对hbase和hdfs进行健康检查，及剩余hdfs容量告警，简单易用
#/bin/bash
bin=`dirname $0`
bin=`cd $bin;pwd`
STATE_OK=0
STATE_WARNING=1
STATE_CRITICAL=2
STATE_UNKNOWN=3
STATE_DEPENDENT=4
source /etc/profile
DFS_REMAINING_WARNING=15
DFS_REMAINING_CRITICAL=5
ABNORMAL_QUERY="INCONSISTENT|CORRUPT|FAILED|Exception"
HADOOP_WEB_INTERFACE=hadoop的Namenode对外接口ip
# hbck and fsck report
output=/data/logs/cluster-status
$HBASE_HOME/bin/hbase hbck >> $output
$HADOOP_HOME/bin/hadoop fsck /hbase >> $output
# check report
count=`egrep -c "$ABNORMAL_QUERY" $output`
if [ $count -eq 0 ]; then
echo "[OK] Cluster is healthy." >> $output
else
echo "[ABNORMAL] Cluster is abnormal!" >> $output
# Get the last matching entry in the report file
last_entry=`egrep "$ABNORMAL_QUERY" $output | tail -1`
echo "($count) $last_entry"
exit $STATE_CRITICAL
fi
# Check RegionServer Status
dead_region_servers=`curl -s http://${HADOOP_WEB_INTERFACE}:60010/master-status | grep "Dead Region Servers" -A 500 | grep "Regions in Transition" -B 500 | egrep -o ‘target="_blank">.*</a>‘ | awk -F">" ‘{print $2}‘ | awk -F"<" ‘{print $1}‘`
if [ -z $dead_region_servers ];then
echo "[OK] All RegionServers is healthy."
echo "[OK] All RegionServers is healthy." >> $output
else
echo "[ABNORMAL] the dead regionserver list:" >> $output
echo $dead_region_servers >> $output
exit $STATE_CRITICAL
fi
# HDFS usage
dfs_remaining=`curl -s http://${HADOOP_WEB_INTERFACE}:50070/dfshealth.jsp |egrep -o "DFS Remaining%.*%" | egrep -o "[0-9]*\.[0-9]*"`
dfs_remaining_word="DFS Remaining%: ${dfs_remaining}%"
echo "$dfs_remaining_word" >> $output
# check HDFS usage
dfs_remaining=`echo $dfs_remaining | awk -F ‘.‘ ‘{print $1}‘`
if [ $dfs_remaining -lt $DFS_REMAINING_CRITICAL ]; then
echo "Low DFS space. $dfs_remaining_word"
exit_status=$STATE_CRITICAL
elif [ $dfs_remaining -lt $DFS_REMAINING_WARNING ]; then
echo "Low DFS space. $dfs_remaining_word"
exit_status=$STATE_WARNING
else
echo "HBase check OK - DFS and HBase healthy.
$dfs_remaining_word"
exit_status=$STATE_OK
fi
exit $exit_status