Introduction to Apache Spark

Apache Spark is an open-source cluster-computing framework. Apache Spark has as its architectural foundation the resilient distributed dataset (RDD), a read-only multi-set of data items distributed over a cluster of machines, that is maintained in a fault-tolerant way. In Spark 1.x, the RDD was the primary application programming interface (API), but as of Spark 2.x use of the Dataset API is encouraged even though the RDD API is not deprecated. The RDD technology still underlies the Dataset API.

Spark and its RDDs were developed in 2012 in response to limitations in the Map-reduce cluster computing paradigm, which forces a particular linear data-flow structure on distributed programs: Map-reduce programs read input data from disk, map a function across the data, reduce the results of the map, and store reduction results on disk. Spark’s RDDs function as a working set for distributed programs that offers a (deliberately) restricted form of distributed shared memory.

Spark facilitates the implementation of both iterative algorithms, that visit their data set multiple times in a loop, and interactive/exploratory data analysis, i.e., the repeated database-style querying of data. The latency of such applications may be reduced by several orders of magnitude compared to a Map-reduce implementation (as was common in Apache Hadoop stacks). Among the class of iterative algorithms are the training algorithms for machine learning systems, which formed the initial impetus for developing Apache Spark.

Apache Spark requires a cluster manager and a distributed storage system. For cluster management, Spark supports standalone (native Spark cluster), Hadoop YARN, or Apache Mesos. For distributed storage, Spark can interface with a wide variety, including Hadoop Distributed File System (HDFS), MapR File System (MapR-FS), Cassandra, OpenStack Swift, Amazon S3, Kudu, or a custom solution can be implemented. Spark also supports a pseudo-distributed local mode, usually used only for development or testing purposes, where distributed storage is not required and the local file system can be used instead; in such a scenario, Spark is run on a single machine with one executor per CPU core.

Since its release, Apache Spark, the unified analytics engine, has seen rapid adoption by enterprises across a wide range of industries. Internet powerhouses such as Netflix, Yahoo, and eBay have deployed Spark at massive scale, collectively processing multiple petabytes of data on clusters of over 8,000 nodes. It has quickly become the largest open source community in big data, with over 1000 contributors from 250+ organizations.

Spark Core is the foundation of the overall project. It provides distributed task dispatching, scheduling, and basic I/O functionalities, exposed through an application programming interface (for Java, Python, Scala, and R) centered on the RDD abstraction (the Java API is available for other JVM languages, but is also usable for some other non-JVM languages, such as Julia, that can connect to the JVM). This interface mirrors a functional/higher-order model of programming: a “driver” program invokes parallel operations such as map, filter or reduce on an RDD by passing a function to Spark, which then schedules the function’s execution in parallel on the cluster. These operations, and additional ones such as joins, take RDDs as input and produce new RDDs. RDDs are immutable and their operations are lazy; fault-tolerance is achieved by keeping track of the “lineage” of each RDD (the sequence of operations that produced it) so that it can be reconstructed in the case of data loss. RDDs can contain any type of Python, Java, or Scala objects.

The above is a brief about Apache Spark. Watch this space for more updates on the latest trends in Technology.

Leave a Reply

Your email address will not be published. Required fields are marked *