Apache Spark is an open source big data processing framework that is faster than Hadoop, easier to use, and supports more types of analytics. It provides high-level APIs, can run computations directly in memory for faster performance, and supports a variety of data processing workloads including SQL queries, streaming data, machine learning, and graph processing. Spark also has a large ecosystem of additional libraries and tools that expand its capabilities.
2. WHAT IS SPARK
Apache Spark is an open source big data
processing framework built around speed, ease
of use, and sophisticated analytics. It was
originally developed in 2009 in UC Berkeley’s
AMPLab, and open sourced in 2010 as an
Apache project.
B I G D A T A W O R K G R O U P . I R
3. WHAT IS SPARK
Advantages: In Memory
Spark enables applications in Hadoop clusters to run up to 100
times faster in memory and 10 times faster even when running
on disk.
B I G D A T A W O R K G R O U P . I R
4. WHAT IS SPARK
Advantages: Generic API
Spark lets you quickly write applications in Java, Scala, or
Python. It comes with a built-in set of over 80 high-level
operators. And you can use it interactively to query data within
the shell.
B I G D A T A W O R K G R O U P . I R
5. WHAT IS SPARK
Advantages: Many Applications
Spark gives us a comprehensive, unified framework to manage
big data processing requirements with a variety of data sets
that are diverse in nature (text data, graph data etc) as well as
the source of data (batch v. real-time streaming data).
B I G D A T A W O R K G R O U P . I R
6. WHAT IS SPARK
Advantages: Many Applications
In addition to Map and Reduce operations, it supports SQL
queries, streaming data, machine learning and graph data
processing. Developers can use these capabilities stand-alone
or combine them to run in a single data pipeline use case.
B I G D A T A W O R K G R O U P . I R
7. HADOOP AND SPARK
Hadoop Spark
Map & Reduce -> suitable for on-
pass computations
multi-step data pipelines using
directed acyclic graph (DAG)
pattern.
Clusters are hard to set up and
manage
supports in-memory data sharing
across DAGs.
need to integrate with Mahout
(Machine Learning) and Storm
(Streaming data processing)
Spark as an alternative to Hadoop
MapReduce
B I G D A T A W O R K G R O U P . I R
8. SPARK FEATURES
Less expensive shuffles in the data processing. With capabilities like in-
memory data storage
Lazy evaluation of big data queries, which helps with optimization of the
steps in data processing workflows.
Higher level API to improve developer productivity and a consistent
architect model for big data solutions.
B I G D A T A W O R K G R O U P . I R
9. SPARK FEATURES
Spark holds intermediate results in memory rather than writing them to
disk
Spark can be used for processing datasets that larger than the aggregate
memory in a cluster.
B I G D A T A W O R K G R O U P . I R
10. SPARK ECOSYSTEM
Spark Streaming
micro batch style of computing and processing.(DStream)
Spark SQL
JDBC API, SQL like queries, ETL
Spark Mlib
including classification, regression, clustering, collaborative filtering,
dimensionality reduction, as well as underlying optimization primitives
B I G D A T A W O R K G R O U P . I R
11. SPARK ECOSYSTEM
Spark GraphX
GraphX extends the Spark RDD by introducing the
Resilient Distributed Property Graph
Set of fundamental operators (e.g., subgraph,
joinVertices, and aggregateMessages)
B I G D A T A W O R K G R O U P . I R
12. SPARK ECOSYSTEM
BlinkDB
trade-off query accuracy for response time.
Tachyon
Caches working set files in memory
Spark Cassandra Connector
access data stored in a Cassandra database
SparkR
B I G D A T A W O R K G R O U P . I R
15. RESILIENT DISTRIBUTED DATASETS
Fault tolerance because an RDD know how to recreate and re-compute the
datasets.
RDDs are immutable.
B I G D A T A W O R K G R O U P . I R