OceanTEA: Exploring Ocean-Derived Climate Data Using Microservices
Arne Johanson, Sascha Flögel, Wolf-Christian Dullo, Wilhelm Hasselbring
Arne Johanson, Wilhelm Hasselbring Software Engineering Group
Kiel University
{arj,wha}@informatik.uni-kiel.de www.se.informatik.uni-kiel.de
www.futureocean.org
Contact
Schematic view of a MoLab configuration
1. Context
Autonomous ocean observation systems, such as the modular ocean laboratory MoLab developed at GEOMAR, produce an increasing amount of time series data. The software tool OceanTEA leverages modern web technology to support scientists in interactively exploring and analyzing such high-dimensional datasets.
The microservice software architecture of OceanTEA
<<microservice>>
Spatial Analysis
<<web browser>>
Oceanographic Time Series Exploration and Analysis Client
<<executionEnvironment>>
NodeJS (REST Wrapper)
<<executionEnvironment>>
R
<<database>>
RDS Data Storage
<<executionEnvironment>>
JavaScript
<<microservice>>
Time Series Pattern Discovery
<<executionEnvironment>>
Python
<<database>>
Netflix Atlas API Gateway
<<microservice>>
Univariate Time Series Management
<<microservice>>
Multivariate Time Series Management
<<executionEnvironment>>
NodeJS
<<database>>
JSON Data Storage
<<executionEnvironment>>
Python
<<database>>
Pickle Data Storage
<<database>>
NumPy Array Storage
<<microservice>>
Time Series Conversion (TEOS-10)
<<executionEnvironment>>
NodeJS (REST Wrapper)
<<executionEnvironment>>
Hosted C Environment
<<microservice>>
User Authentication
<<service>>
Google Maps
Data Exchange
REST REST REST
REST
HTTP, REST REST
3. Microservice Architecture
The implementation of OceanTEA is partitioned into so-called microservices, which are small, self-contained applications that can be deployed independently and each have a single functional responsibility.
Optimal implementation and storage technologies for each microservice
Scales seamlessly from desktop computers to cloud computing infrastructure
•
•
The data exploration view of OceanTEA
2. OceanTEA
Open-source tool to support
interactive data visualization spatial analysis
temporal pattern exploration
for both univariate and multivariate time series.
Try the live demo of OceanTEA:
github.com/a-johanson/oceantea
•
•
•
