HP Hadoop Platform
High Performance – High Throughput and Low Latency
HP Hadoop Platform
Reliable Scalable Cost Effective
• Execute High through put Transformations
• High Reliability and fault tolerance
• MR Engine for High Throughput
• TEZ for faster response
• Hive - High Productivity for SQL Developers
• Pig – Flow visualization makes understanding easy
• Pig and Hive Integration using HCatalog
HP Hadoop Platform
Hive/Pig– High Throughput Queries
• Presto is in-memory based- stores query results in memory and subsequent
operations on query result are faster.
• Presto works on top of HDFS and YARN
• Leverages Hive Meta store
• Offers flexibility in deployment- Can be deployed on few nodes in cluster
• HDP Certified release available
HP Hadoop Platform
Presto – Low Latency Queries
• HDFS Storage– Data Locality. Suited for Lambda architecture
• On top of HBase/Big Table columnar storage
• Consistency in CAP theorem
• Linear Scalability
• HTTP API for pulling Readings.
• Write with millisecond precision
• Integrates with Visualization tools
OpenTSDB – Time Series Solution
HP Hadoop Platform
• In Memory computation engine
• 100x faster than MapReduce.
• Ships with ML Lib – Machine Learning Library
• Usable with Scala/Python/R/Java
• Integrates with Kafka, Hive SQL, OpenTSDb
• GraphX/Graph Frame – Graph Query Capabilities
HP Hadoop Platform
Spark – Speed up Computation and ML
• Supported on Hortonworks Platform
• Based on Lucene Search engine
• HDFS Storage – Scalable storage
• Solr Cloud - Distributed architecture
• Near Real-Time Indexing
• Phonetic Matching supported
HP Hadoop Platform
Solr Indexing – Scalable Search

Hp hadoop platform

  • 1.
    HP Hadoop Platform HighPerformance – High Throughput and Low Latency
  • 2.
    HP Hadoop Platform ReliableScalable Cost Effective
  • 3.
    • Execute Highthrough put Transformations • High Reliability and fault tolerance • MR Engine for High Throughput • TEZ for faster response • Hive - High Productivity for SQL Developers • Pig – Flow visualization makes understanding easy • Pig and Hive Integration using HCatalog HP Hadoop Platform Hive/Pig– High Throughput Queries
  • 4.
    • Presto isin-memory based- stores query results in memory and subsequent operations on query result are faster. • Presto works on top of HDFS and YARN • Leverages Hive Meta store • Offers flexibility in deployment- Can be deployed on few nodes in cluster • HDP Certified release available HP Hadoop Platform Presto – Low Latency Queries
  • 5.
    • HDFS Storage–Data Locality. Suited for Lambda architecture • On top of HBase/Big Table columnar storage • Consistency in CAP theorem • Linear Scalability • HTTP API for pulling Readings. • Write with millisecond precision • Integrates with Visualization tools OpenTSDB – Time Series Solution HP Hadoop Platform
  • 6.
    • In Memorycomputation engine • 100x faster than MapReduce. • Ships with ML Lib – Machine Learning Library • Usable with Scala/Python/R/Java • Integrates with Kafka, Hive SQL, OpenTSDb • GraphX/Graph Frame – Graph Query Capabilities HP Hadoop Platform Spark – Speed up Computation and ML
  • 7.
    • Supported onHortonworks Platform • Based on Lucene Search engine • HDFS Storage – Scalable storage • Solr Cloud - Distributed architecture • Near Real-Time Indexing • Phonetic Matching supported HP Hadoop Platform Solr Indexing – Scalable Search