/pipelines

Pipelines for data processing (GBIF and LivingAtlases)

Primary LanguageJavaApache License 2.0Apache-2.0

DEV - Build Status Coverage

Table of Contents

About the project

REMEMBER, YOU HAVE TO USE JAVA VERSION 8

Pipelines for data processing and indexing of biodiversity data

Status: IN PRODUCTION

Vision: Consistent data processing pipelines (parsing, interpretation and quality flagging) for use in GBIF, the Living Atlases project and beyond. Built to scale from laptop to GBIF volumes. Deployable on JVM, Spark, Google Cloud, .

Architecture

The project provides vanilla JVM-based parsing and interpretation libraries, and pipelines for indexing into SOLR and ElasticSearch, built using Apache Beam.

Apache Beam provides a high level abstraction framework ideal for this purpose with the ability to deploy across target environments (e.g. Spark, local JVM) and with many built in connectors for e.g. HBase, SOLR, ElasticSearch etc.

Ingress

Ingress is from Darwin Core Archive (zip files of one or more delimited text files) or ABCD Archives (compressed XML) only[1]. During ingress data is converted from its native format and stored as Avro files containing Darwin Core compliant data.

This is depicted below:

Ingress

Avro is chosen as a storage and interchange format exclusively in this project because a) it is splittable with each split compressed independently, b) it holds the data schema with the data, c) is well supported in the Hadoop ecosystem (e.g. Hive, Spark) and many other tools (e.g. Google Cloud) d) is very robust in serialization and e) reduces boiler plate code thanks to schema to code generators. Darwin Core Archives and JSON for example do not exhibit all these traits.

[1] Other protocols (e.g. DiGIR, TAPIR) are supported by GBIF but are converted by crawling code upstream of this project.

Interpretation

During interpretation the verbatim data is parsed, normalised and tested against quality control procedures.

To enable extensibility data is interpreted into separate avro files where a separate file per category of information is used. Many interpretations such as date / time formatting are common to all deployments, but not all. For example, in the GBIF.org deployment the scientific identification is normalised to the GBIF backbone taxonomy and stored in /interpreted/taxonomy/interpreted*.avro which might not be applicable to everyone. Separating categories allows for extensibility for custom deployments in a reasonably modular fashion.

Interpretation is depicted below:

Ingress

Note that all pipelines are designed and tested to run with the DirectRunner and the SparkRunner at a minimum. This allows the decision to be taken at runtime to e.g. opt to interpret a small dataset in the local JVM without needing to use cluster resources for small tasks.

It is a design decision to ensure that all the underlying parsers are as reusable as possible for other projects with careful consideration to not bring in dependencies such as Beam or Hadoop.

Indexing

Initial implementations will be available for both SOLR and for ElasticSearch to allow for evaluation of both at GBIF. During indexing the categories of interpreted information of use are merged and loaded into the search indexes:

Ingress

Note that GBIF target 10,000 records/sec per node indexing speed (i.e. 100,000 records/sec on current production cluster). This will allow simplified disaster recovery and rapid deployment and of new features.

Structure

The project is structured as:

  • .buildSrc - Tools for building the project
  • docs - Documents related to the project
  • examples - Examples of using project API and base classes
    • dwca-to-elasticsearch - Example how use pipelines, interprets DWCA into ES index
    • transform - Transform example demonstrates how to create Apache Beam pipeline, create the new transformation and use it together with GBIF transforms and core classes
    • metrics - The example demonstrates how to create and send Apache Beam SparkRunner metrics to ELK and use the result for Kibana dashboards
  • gbif - GBIF main module
    • coordinator - The main module which controls the interpretation process, through RabbitMQ
    • diagnostics - Internal tool to fix GBIF identifiers collisions
    • keygen - The library to generate GBIF identifiers, to support backward compatibility the codebase (with minimum changes) was copied from the occurrence/occurrence-persistence project
    • pipelines - Main pipelines module
      • export-gbif-hbase - The pipeline to export the verbatim data from the GBIF HBase tables and save as ExtendedRecord avro files
      • ingest-gbif-beam - Main GBIF pipelines for ingestion of biodiversity data
      • ingest-gbif-fragmenter - Writes raw archive's data to HBase store
      • ingest-gbif-java - Main GBIF pipelines for ingestion of biodiversity data, Java version
    • validator - Main validator module
  • livingatlas - Living atlas main module
  • sdks - Main module contains common classes, such as data models, data format interpretations, parsers, web services clients etc.
    • beam-common - Classes and API for using with Apache Beam
    • beam-transforms - Transformations for ingestion of biodiversity data
    • core - Main API classes, such as data interpretations, converters, DwCA reader etc.
    • models - Data models represented in Avro binary format, generated from Avro schemas
    • variables - Simple static string variables
  • tools - Module for different independent tools

How to build the project

The project uses Apache Maven for building. The project contains a Maven wrapper script for Linux and MacOS systems, you just can run the build.sh script:

./build.sh

Please read the Apache Maven how-to.

Contributions

We welcome contributions! Please see the guide to see how you can best do this.