Hadoop distributed file system
Download as PPTX, PDF3 likes2,594 views
HDFS is a distributed file system designed to run on commodity hardware. It provides high-performance access to big data across Hadoop clusters and supports big data analytics applications in a low-cost manner. The NameNode stores metadata and manages the file system namespace, while DataNodes store file data in blocks and handle replication for fault tolerance. Clients interact with the NameNode for file operations like writing blocks to DataNodes for storage and reading file blocks.
Related topics:1 of 11
Downloaded 120 times
Hadoop distributed file system
- 1. Hadoop Distributed File System Submitted By: Anshul Bhatnagar Amit Sharma Abhishek Pareek (VII Sem CS-A)
- 2. What is HDFS ? Distributed File System. Designed to run on commodity hardware. Tool for managing pools of big data. High-Performance access to data across Hadoop clusters. Supports big data analytics applications. Low-Cost.
- 3. HDFS ARCHITECTURE Advertising OF Storage. Namenode DatanodeDatanodeDatanodeDatanodeDatanode
- 4. What Is NameNode? Centerpiece of an HDFS file system. Keeps the directory tree of all files in the file system. It does not store the data of these files itself. It contains Meta data of the files stored in the system. If a name node gets down , the system goes offline.
- 5. What is Data Node? Advertise storage. Stores data in the HDFS. There can be many data nodes. Replication is done on data nodes. If a data node gets offline it won’t affect the cluster.
- 6. HDFS Properties Stripping of a file (Making Blocks). Default Block Size is 64 Mb. Every block is replicated three times (By Default). Sending of heartbeat (Keep alive message) in every 3 seconds (Default). Fault Tolerance.
- 7. HDFS Quotas Name Quotas It is hard limit on number of files and directory names in the tree rooted at the directory. Space Quotas It is hard limit on number of bytes used by files in the tree rooted at the directory.
- 8. Client & HDFS Cluster. There are two major operations done by a client: Storing of a file. Reading of a file.
- 9. Storing a file in HDFS. Namenode DatanodeDatanodeDatanodeDatanode Client Block1 Block2 Block3 Replication Replication
- 10. Any Query?
- 11. Thank You