HADOOP-4959. Correct system metric collection from top on Redhat 5.1. Contributed by Eric Yang

git-svn-id: https://svn.apache.org/repos/asf/hadoop/core/trunk@742793 13f79535-47bb-0310-9956-ffa450edef68

CHANGES.txt

@@ -117,6 +117,9 @@ Trunk (unreleased changes)
     HADOOP-4859. Distinguish daily rolling output dir by adding a timestamp.
     (Jerome Boulon via cdouglas)
 
+    HADOOP-4959. Correct system metric collection from top on Redhat 5.1. (Eric
+    Yang via cdouglas)
+
 Release 0.20.0 - Unreleased
 
   INCOMPATIBLE CHANGES

src/contrib/chukwa/bin/systemDataLoader.sh

@@ -57,6 +57,12 @@ fi
 
 echo -n "Starting System Data Loader..."
 
+export SAR="sar -q -r -n ALL 55"
+RELEASE=`lsb_release -r | cut -b 10-`
+if [ "X${RELEASE}" = "X4" ]; then 
+    export SAR="sar -q -r -n FULL 55"
+fi
+
 #test=`grep -q SysLog ${CHUKWA_HOME}/var/chukwa_checkpoint*`
 #if [ "X${test}"="X1" ]; then
 #  echo "add org.apache.hadoop.chukwa.datacollection.adaptor.filetailer.CharFileTailingAdaptorUTF8NewLineEscaped SysLog 0 /var/log/messages 0" | nc localhost 9093 >&/dev/null & disown -h 
@@ -73,7 +79,7 @@ if [ -f $pidFile ]; then
 fi
 
 if [ ${EXISTS} -lt 1 ]; then
-    ${JAVA_HOME}/bin/java $JVM_OPTS -DPERIOD=60 -DCHUKWA_HOME=${CHUKWA_HOME} -DCHUKWA_CONF_DIR=${CHUKWA_CONF_DIR} -DCHUKWA_LOG_DIR=${CHUKWA_LOG_DIR} -DRECORD_TYPE=Sar -Dlog4j.configuration=system-data-loader.properties -classpath ${CLASSPATH}:${CHUKWA_CORE}:${HADOOP_JAR}:${COMMON}:${TOOLS}:${CHUKWA_CONF_DIR} org.apache.hadoop.chukwa.inputtools.plugin.metrics.Exec sar -q -r -n FULL 55 &
+    ${JAVA_HOME}/bin/java $JVM_OPTS -DPERIOD=60 -DCHUKWA_HOME=${CHUKWA_HOME} -DCHUKWA_CONF_DIR=${CHUKWA_CONF_DIR} -DCHUKWA_LOG_DIR=${CHUKWA_LOG_DIR} -DRECORD_TYPE=Sar -Dlog4j.configuration=system-data-loader.properties -classpath ${CLASSPATH}:${CHUKWA_CORE}:${HADOOP_JAR}:${COMMON}:${TOOLS}:${CHUKWA_CONF_DIR} org.apache.hadoop.chukwa.inputtools.plugin.metrics.Exec $SAR &
 fi
 
 EXISTS=0
@@ -87,7 +93,7 @@ if [ -f $pidFile ]; then
 fi
 
 if [ ${EXISTS} -lt 1 ]; then
-  ${JAVA_HOME}/bin/java $JVM_OPTS -DPERIOD=60 -DCHUKWA_HOME=${CHUKWA_HOME} -DCHUKWA_CONF_DIR=${CHUKWA_CONF_DIR} -DCHUKWA_LOG_DIR=${CHUKWA_LOG_DIR} -DRECORD_TYPE=Iostat -Dlog4j.configuration=system-data-loader.properties -classpath ${CLASSPATH}:${CHUKWA_CORE}:${HADOOP_JAR}:${COMMON}:${TOOLS}:${CHUKWA_CONF_DIR} org.apache.hadoop.chukwa.inputtools.plugin.metrics.Exec iostat -x 55 2 &
+  ${JAVA_HOME}/bin/java $JVM_OPTS -DPERIOD=60 -DCHUKWA_HOME=${CHUKWA_HOME} -DCHUKWA_CONF_DIR=${CHUKWA_CONF_DIR} -DCHUKWA_LOG_DIR=${CHUKWA_LOG_DIR} -DRECORD_TYPE=Iostat -Dlog4j.configuration=system-data-loader.properties -classpath ${CLASSPATH}:${CHUKWA_CORE}:${HADOOP_JAR}:${COMMON}:${TOOLS}:${CHUKWA_CONF_DIR} org.apache.hadoop.chukwa.inputtools.plugin.metrics.Exec iostat -x -k 55 2 &
 fi
 
 EXISTS=0

src/contrib/chukwa/src/java/org/apache/hadoop/chukwa/extraction/demux/processor/mapper/Top.java

@@ -126,7 +126,7 @@ public class Top extends AbstractProcessor
 			record.add("tasks_stopped",matcher.group(4));
 			record.add("tasks_zombie",matcher.group(5));
 		}
-		p = Pattern.compile("Cpu\\(s\\):\\s+(.*?)% us,\\s+(.*?)% sy,\\s+(.*?)% ni,\\s+(.*?)% id,\\s+(.*?)% wa,\\s+(.*?)% hi,\\s+(.*?)% si");
+		p = Pattern.compile("Cpu\\(s\\):\\s*(.*?)%\\s*us,\\s*(.*?)%\\s*sy,\\s*(.*?)%\\s*ni,\\s*(.*?)%\\s*id,\\s*(.*?)%\\s*wa,\\s*(.*?)%\\s*hi,\\s*(.*?)%\\s*si");
 		matcher = p.matcher(headers[2]);
 		if(matcher.find()) {
 			record.add("cpu_user%",matcher.group(1));
@@ -162,4 +162,4 @@ public class Top extends AbstractProcessor
 	public String getDataType() {
 		return recordType;
 	}
-}
+}