Page 1
HadoopEdit by Cassell Hsu
2013.04.19
Page 2
Hadoop
Master
Slave
Slave
Slave
Page 3
Hadoop•Master•NameNode•JobTracker•SecondaryNameNode
Page 4
NameNode•Where is NameNode?•Master•HDFS(Hadoop Distributed File
System)•What is NameNode?•資料之位置資訊 •資料之屬性
??
Page 5
NameNode
•位置資訊 ?•所有資料皆存放在 – DataNode
Page 6
DataNode•What is DataNode•存放資料
•Where is DataNode•HDFS•Slaves (and Master)
Page 7
UserUser
DataNodeDataNode
NameNodeNameNode
DataNodeDataNode
DataNodeDataNode
128Mb
B64Mb
A64Mb
Check hdfs-site.xml
B64Mb
A64Mb
Page 8
Hadoop
DataNodeDataNodeNameNodeNameNode
MasterDataNodeDataNode
DataNodeDataNode
DataNodeDataNode
Slaves
Page 9
Hadoop•Master•NameNode•JobTracker•SecondaryNameNode
Page 10
JobTracker•What is JobTracker?•排程工作
•Where is JobTracker?•Master
誰來工作?
Page 11
JobTracker & TaskTrackerJobTracker TaskTracker
Where Master Slaves
What 排程工作 執行工作
Page 12
Hadoop
DataNodeDataNodeNameNodeNameNode
MasterDataNodeDataNode
Slaves
JobTrackerJobTracker
TaskTrackerTaskTracker
Page 13
Hadoop•Master•NameNode•JobTracker•SecondaryNameNode
Page 14
SecondaryNameNode
•What is SecondaryNameNode?•NameNode發生錯誤時補救
•Where is SecondaryNameNode?•Master
Page 15
Hadoop
DataNodeDataNodeNameNodeNameNode
MasterDataNodeDataNode
Slaves
JobTrackerJobTracker
TaskTrackerTaskTracker
SecondaryNodeSecondaryNode
Page 16
MapReduce
User Master
Slave2
Slave1
A
A1
A2Task
NameNode
Task
Result1Result2
ReduceFinal
ResultHDFS
Page 17
MapReduce
•檔案切割•Hadoop上區塊切割•程式指定