Hadoop training in bangalore-kellytechnologies
- 1. Presented By
- 2. Framework for running applications on large clusters of commodity hardware Scale: petabytes of data on thousands of nodes Include Storage: HDFS Processing: MapReduce Support the Map/Reduce programming model Requirements Economy: use cluster of comodity computers Easy to use Users: no need to deal with the complexity of distributed computing Reliable: can handle node failures automatically http://www.kellytechno.com
- 3. Implemented in Java Apache Top Level Project http://hadoop.apache.org/core/ Core (15 Committers) HDFS MapReduce Community of contributors is growing Though mostly Yahoo for HDFS and MapReduce You can contribute too! http://www.kellytechno.com
- 4. Commodity HW Add inexpensive servers Storage servers and their disks are not assumed to be highly reliable and available Use replication across servers to deal with unreliable storage/servers Metadata-data separation - simple design Namenode maintains metadata Datanodes manage storage Slightly Restricted file semantics Focus is mostly sequential access Single writers No file locking features Support for moving computation close to data Servers have 2 purposes: data storage and computation Single ‘storage + compute’ cluster vs. Separate clusters http://www.kellytechno.com
- 5. Data Data data data data data Data data data data data Data data data data data Data data data data data Data data data data data Data data data data data Data data data data data Data data data data data Data data data data data Data data data data data Data data data data data Data data data data data Results Data data data data Data data data data Data data data data Data data data data Data data data data Data data data data Data data data data Data data data data Data data data data Hadoop Cluster DFS Block 1 DFS Block 1 DFS Block 2 DFS Block 2 DFS Block 2 DFS Block 1 DFS Block 3 DFS Block 3 DFS Block 3 MAP MAP MAP Reduce http://www.kellytechno.com
- 6. Data is organized into files and directories Files are divided into uniform sized blocks and distributed across cluster nodes Blocks are replicated to handle hardware failure Filesystem keeps checksums of data for corruption detection and recovery HDFS exposes block placement so that computation can be migrated to data http://www.kellytechno.com
- 7. NameNode(Filename, replicationFactor, block-ids, …) /users/user1/data/part-0, r:2, {1,3}, … /users/user1/data/part-1, r:3, {2,4,5}, … Datanodes 1 1 3 3 2 2 2 4 4 4 55 5 http://www.kellytechno.com
- 8. Master-Slave architecture DFS Master “Namenode” Manages the filesystem namespace Maintain file name to list blocks + location mapping Manages block allocation/replication Checkpoints namespace and journals namespace changes for reliability Control access to namespace DFS Slaves “Datanodes” handle block storage Stores blocks using the underlying OS’s files Clients access the blocks directly from datanodes Periodically sends block reports to Namenode Periodically check block integrity http://www.kellytechno.com
- 9. Read Metadata ops Client Metadata (Name, #replicas, …): /users/foo/data, 3, … Namenode Client Datanodes Rack 1 Rack 2 Replication Block ops Datanodes Write Blocks http://www.kellytechno.com
- 10. A file’s replication factor can be set per file (default 3) Block placement is rack aware Guarantee placement on two racks 1st replica is on local node, 2rd/3rd replicas are on remote rack Avoid hot spots: balance I/O traffic Writes are pipelined to block replicas Minimize bandwidth usage Overlapping disk writes and network writes Reads are from the nearest replica Block under-replication & over-replication is detected by Namenode Balancer application rebalances blocks to balance DN utilization http://www.kellytechno.com
- 11. Scale cluster size Scale number of clients Scale namespace size (total number of files, amount of data) Possible solutions Multiple namenodes Read-only secondary namenode Separate cluster management and namespace management Dynamic Partition namespace Mounting http://www.kellytechno.com
- 12. Map/Reduce is a programming model for efficient distributed computing It works like a Unix pipeline: cat input | grep | sort | uniq -c | cat > output Input | Map | Shuffle & Sort | Reduce | Output A simple model but good for a lot of applications Log processing Web index building http://www.kellytechno.com
- 14. Mapper Input: value: lines of text of input Output: key: word, value: 1 Reducer Input: key: word, value: set of counts Output: key: word, value: sum Launching program Defines the job Submits job to cluster http://www.kellytechno.com
- 15. Fine grained Map and Reduce tasks Improved load balancing Faster recovery from failed tasks Automatic re-execution on failure In a large cluster, some nodes are always slow or flaky Framework re-executes failed tasks Locality optimizations With large data, bandwidth to data is a problem Map-Reduce + HDFS is a very effective solution Map-Reduce queries HDFS for locations of input data Map tasks are scheduled close to the inputs when possible http://www.kellytechno.com
- 16. • Hadoop Wiki – Introduction • http://hadoop.apache.org/core/ – Getting Started • http://wiki.apache.org/hadoop/GettingStartedWithHadoop – Map/Reduce Overview • http://wiki.apache.org/hadoop/HadoopMapReduce – DFS • http://hadoop.apache.org/core/docs/current/hdfs_design.html • Javadoc – http://hadoop.apache.org/core/docs/current/api/index.html http://www.kellytechno.com
Editor's Notes
- #9: 60M objects on 16G machine (e.g. 20M files with 2 blocks each)