image/svg+xml
REST API on top • List locations, mapsets, vector-, raster- and spatio-temporal data as resources • Enable usage of GRASS modules • User Management - Groups for mapset access - Roles e.g. for pixel limits, number of processes and calculation duration limits • Mapset locking to avoid conflicts • GRASS DB management
GRASS GIS in the cloud: actinia geoprocessing
User defined processing • ephemeral• persistent Create custom process chains bychaining the processing toolsprovided by actinia:- GRASS modules,- importer/exporter,- python scripts,- cmd (GDAL, PROJ, snappy),- ...
"bring algorithm to the data" • via predefined functionalities • via Importer (/Exporter) • e.g. Sentinel data (S-1 and S-2) processing • Access: - Download from ESA API hub - from local storage (as provided by DIAS platforms) - Amazon AWS*- Google Cloud Storages*- ... *advantage of single band access
outlook• implement module / process self-description• run predefined processes via generic endpoint (process chain CRUD)• process across GRASS locations (reprojection)• refactoring (more transparency, clearer file structure, split classes, use python 3.7, typing, swagger3)• relocate Amazon / Googlecloud connection to actinia-plugins• enable use of dynamic process chain
more features• export to GCS + S3 • ace (actinia command execution via GRASS GIS)• advanced usage with GRASS addons• openEO support - openeo-grassgis-driver converts processing requests (process graphs) into actinia jobs. The openEO core API (https://openeo.org; H2020 project) is a standardized interface supporting multiple backends.• actinia-algebra (cli-tool, proof-of-concept) to calculate massively in parallel. It calls multiple actinia endpoints (create mapsets, processing_async, merge mapsets, create strds, delete mapsets). It splits a calculation into individual processes, processes everything in parallel, and puts the result together.
outlook• implement module / process self-description• run predefined processes via generic endpoint (process chain CRUD)• process across GRASS locations (reprojection)• refactoring (more transparency, clearer file structure, split classes, use python 3.7, typing, swagger3)• relocate Amazon / Googlecloud connection to actinia-plugins• enable use of dynamic process chain
join on github:* https://github.com/mundialis/actinia_core* https://github.com/mundialis/actinia-gdi* https://github.com/mundialis/actinia_satellite_plugin* https://github.com/mundialis/actinia_statistic_plugin get started:* https://actinia.mundialis.de/api_docs
Presentation created with Inkscape and Sozi - https://sozi.baierouge.fr / https://github.com/senshu/SoziIcons designed by Freepik, Smashicons and surang from Flaticon