Post by Admin on Mar 14, 2014 2:28:07 GMT
HDFS CLUSTER
•A client access the file system on behalf of user by communicating with the Name Node.
•The client access files system to a portable operating system Interface (POSIX). So, the user code does not to know about the Name Node.
•Assign Blocks to Data nodes.
•Keeps track of live nodes (through heart beats)
•Limitates re-replication in case of data node loss.
•Block metadata is held in memory.
will run out of memory when too many files exist.
•Is a single point of failure in the system
Some solutions exist.
•A client access the file system on behalf of user by communicating with the Name Node.
•The client access files system to a portable operating system Interface (POSIX). So, the user code does not to know about the Name Node.
•Assign Blocks to Data nodes.
•Keeps track of live nodes (through heart beats)
•Limitates re-replication in case of data node loss.
•Block metadata is held in memory.
will run out of memory when too many files exist.
•Is a single point of failure in the system
Some solutions exist.