Apache Hadoop
See Hadoop.
Apache Spark
See Spark Program.
A collection of programs and services that read and write through the data abstraction layer in CDAP.
Application Abstraction
Application abstraction allows the same application to run in multiple environments without modification.
Application Template
An artifact, that with the addition of a configuration file, can be used to create manifestations of applications.
A JAR file containing Java classes and resources required to create and run an Application. Multiple applications can be created from the same artifact.
Refers to the Apache Avro™ project, which is a data serialization system that provides rich data structures and a compact, fast, binary data format.
Batch Pipeline
A type of CDAP Pipeline that runs on a schedule, performing actions on a distinct set of data.
The Cask Data Application Platform; refers to both the platform, and an installed instance of it.
CDAP Application
See Application.
See Command Line Interface.
CDAP Console
CDAP Pipeline
A CDAP application; created from an application template, generally one of the system artifacts shipped with CDAP; defines a source to read from, zero or more transformations or other steps to perform on the data that was read from the source, and one or more sinks to write the transformed data to.
CDAP Pipeline Plugin
A plugin of type BatchSource, RealtimeSource, BatchSink, RealtimeSink, or Transformation, packaged in a JAR file format, for use as a plugin in a CDAP pipeline.
CDAP Sandbox
A version of the Cask Data Application Platform, supplied as a downloadable archive, that runs on a single machine in a single Java Virtual Machine (JVM). It provides all of the CDAP APIs without requiring a Hadoop cluster, using alternative, fully-functional implementations of CDAP features. For example, application containers are implemented as Java threads instead of YARN containers. Formerly known as the Standalone CDAP.
CDAP Studio
A visual editor, part of the CDAP UI, for creating and configuring pipelines. You click and drag sources, transformations, and sinks, and can name and configure the pipelines. It provides an operational view of the resulting pipeline that allows for monitoring of metrics, logs, and other run-time information.
The CDAP UI is a web-based application used to deploy CDAP applications, create pipelines using the CDAP Studio, and query and manage the Cask Data Application Platform instance.
Command Line Interface
The Command Line Interface (CLI) provides methods to interact with CDAP from within a shell, similar to the HBase or bash shells.
A consumer is an entity that is a recipient of events, typically a flowlet. A consumer flowlet is the recipient of a queue. The difference between the total of a producer's events and the consumed events is the pending events.
A directed acyclic graph. Flows are wired together and displayed as a DAG in the CDAP UI.
Data Abstraction
Abstraction of the actual representation of data in storage.
Data Pipeline
A type of pipeline, often not linear in nature and require the performing of complex transformations including forks and joins at the record and feed level. They can be configured to perform various functions at different times, including machine-learning algorithms and custom processing.
Datasets store and retrieve data and are a high-level abstraction of the underlying data storage with generic reusable implementations of common data patterns.
Distributed CDAP
A version of the Cask Data Application Platform, supplied as either Yum .rpm or APT .deb packages, that runs on a Hadoop cluster. Packages are available for Ubuntu 12 and CentOS 6.
Abbreviation for extract, transform, and loading of data.
Refers to the Extract, Transform and Load of data.
Streams and datasets in CDAP can be explored through ad-hoc SQL-like queries. To enable exploration, you must set several properties when creating the stream or dataset, and the files in a dataset must meet certain requirements.
A dataset composed of collections of files in the file system that share some common attributes such as the format and schema, which abstracts from the actual underlying file system interfaces.
Flows are user-implemented real-time stream processors, comprised of one or more flowlets that are wired together into a directed acyclic graph.
A flowlet represents an individual processing node within a flow. flowlets consume data objects from their inputs and execute custom logic on each data object, able to perform data operations as well as emit data objects to the flowlet’s outputs.

Refers to the Apache Hadoop® project, which describes itself as:

"The Apache Hadoop software library is a framework that allows for the distributed processing of large data sets across clusters of computers using simple programming models. It is designed to scale up from single servers to thousands of machines, each offering local computation and storage. Rather than rely on hardware to deliver high-availability, the library itself is designed to detect and handle failures at the application layer, so delivering a highly-available service on top of a cluster of computers, each of which may be prone to failures."

Logical Pipeline
A view of a pipeline composed of sources, sinks, and other plugins, and does not show the underlying technology used to actually manifest and run the pipeline.
MapReduce is a processing model used to process data in batch. MapReduce programs can be written as in a conventional Apache Hadoop system. CDAP datasets can be accessed from MapReduce programs as both input and output.
Master Services
CDAP system services that are run in YARN containers, such as the Transaction Service, Dataset Executor, Log Saver, Metrics Processor, etc.
A namespace is a logical grouping of application, data and its metadata in CDAP. Conceptually, namespaces can be thought of as a partitioning of a CDAP instance. Any application or data (referred to here as an “entity”) can exist independently in multiple namespaces at the same time. The data and metadata of an entity is stored independent of another instance of the same entity in a different namespace. The primary motivation for namespaces in CDAP is to achieve application and data isolation.
Physical Pipeline
A physical pipeline is the manifestation of a logical pipeline as a CDAP application, which is a collection of programs and services that read and write through the data abstraction layer in CDAP.
CDAP provides an easy method of configuring pipelines using a visual editor, called CDAP Studio. You click and drag sources, transformations, and sinks, configuring an pipeline within minutes. It provides an operational view of the resulting pipeline that allows for monitoring of metrics, logs, and other run-time information.
A pipeline is a series of stages—linked usages of individual programs—configured together into an application.
A plugin extends an application by implementing an interface expected by the application. Plugins are packaged in an artifact.
A plugin extends an application template by implementing an interface expected by the application template. Plugins are packaged in an artifact.
A producer is an entity that emits events, such as a stream or a flowlet. Flowlets can at the same time be both consumers and producers.
Real-time Pipeline
A type of CDAP Pipeline that runs continuously, performing actions on a distinct set of data.
Route Config
See route configuration.
Route Configuration
Also known as a route config, a map that allocates requests for a service between different versions of the service.
Secure Key
An identifier or an alias for an entry in Secure Storage. An entry in secure storage can be referenced and retrieved using a Secure Key using programmatic or RESTful APIs.
Secure Storage
Encrypted storage for sensitive data using Secure Keys. CDAP supports File-backed (for CDAP Sandbox) as well as Apache Hadoop KMS-backed (for Distributed CDAP) Secure Storage.
Services can be run in a Cask Data Application Platform (CDAP) application to serve data to external clients. Similar to flows, services run in containers and the number of running service instances can be dynamically scaled. Developers can implement custom services to interface with a legacy system and perform additional processing beyond the CDAP processing paradigms. Examples could include running an IP-to-Geo lookup and serving user-profiles.
Spark is a fast and general processing engine, compatible with Hadoop data, used for in-memory cluster computing. It lets you load large sets of data into memory and query them repeatedly, making it suitable for both iterative and interactive programs. Similar to MapReduce, Spark can access datasets as both input and output. Spark programs in CDAP can be written in either Java or Scala.
Standalone CDAP
See CDAP Sandbox.
Storage Provider
For datasets and streams, a storage provider is the underlying system that CDAP uses for persistence. Examples include HDFS, HBase, and Hive.
The primary means of bringing data from external systems into CDAP in real time; an ordered, time-partitioned sequences of data, usable for real-time collection and consumption of data.
Structured Record
The data format used to exchange events between most of the pre-built CDAP ETL plugins.
Structured Record
A data format, defined in CDAP, that can be used to exchange events between plugins. Used by many of the CDAP pipeline plugins included in CDAP.
System Artifact
An application template, shipped with CDAP, that with the addition of a configuration file, can be used to create manifestations of applications.
Time-partitioned FileSet
A FileSet dataset that uses a timestamp as the partitioning key to split the data into indivividual files. Though it is not required that data in each partition be organized by time, each partition is assigned a logical time. Typically written to in batch mode, at a set time interval.
Timeseries Dataset
A dataset where time is the primary means of how data is organized, and both the data model and the schema that represents the data are optimized for querying and aggregating over time ranges.
A read-only view of a stream, with a specific read format. Read formats consist of a schema and a format (such as CSV, TSV, or Avro, amongst others).
Workers are typically long-running background programs that can be used to execute tasks. Each instance of a worker runs either in its own YARN container (Distributed CDAP mode) or a single thread (CDAP Sandbox or in-memory mode) and the number of instances may be updated via RESTful APIs or the CLI. Datasets can be accessed from inside workers.
A workflow is used to execute a series of MapReduce programs, with an optional schedule to run itself periodically.