This repository is a template for creating new Cirrus tasks, either as Lambda functions or as a Batch job.
Include this task in a workflow by including the Lambda deployment package URL or the Docker image name for Batch.
| Version | Package URL (Lambda) | Image Name (Batch) |
|---|---|---|
| 0.2.0 | https://url/to/zip/file | org/image:tag |
Configuration parameters are available in the Cirrus payload under payload['process']['tasks']['<taskname>']. The following parameters are available:
| Field | Type | Description |
|---|---|---|
| assets | Map<string, ConvertAsset Object> | REQUIRED Dictionary of Asset keys to convert with parameters for each asset |
| drop_assets | [string] | Asset keys to remove from output STAC Item(s) (Default: []) |
Additionally this task creates files that will be uploaded as assets using the parameters supplied in payload['process']['output_options'] (see cirruslib.transfer.upload_item_assets)
Tasks can be run locally as they include a CLI:
$ task.py -h
When run in Cirrus the cirrus subcommand is used when invoking Batch
$ task.py cirrus s3://bucket/prefix/payload.json
While when run locally use the local subcommand, specify a local payload, and optionally provide a working directory
$ task.py local tests/payload.json --workdir tests/
When run locally the handler will not upload assets to s3. It will also not remove the working directory when it completes.
Cirrus is an open-source pipeline for processing geospatial data in AWS. Cirrus was developed by Element 84 originally under a [NASA ACCESS project]. It is released under the Apache License.