New in EUMDAC 2.1.0
- Added automated retry and Data Store throttling management mechanisms
- New CLI capabilities (improved describe, new flags, progress bar and throughput measurement)
- Added support for local Data Tailor instances
- Improved error reporting
Added automated retry and Data Store throttling management mechanisms
The EUMDAC library now includes a requests.py module that will take care of retrying failed requests, unless they fail due to invalid user input, in a transparent way.
This module will also manage cases of Data Store requests being throttled. See Data Store 2.2 introduces throttling for more information on Data Store throttling (although eumdac will now take care of everything!).
New CLI capabilities (improved describe, new flags, progress bar and throughput measurement)
The eumdac describe
command has been updated to show more information, including the available search parameters for each collection and how they map to CLI parameters.
The eumdac describe
command has been updated to show more information, including the available search parameters for each collection and how they map to CLI parameters.
Searches are now more powerful than ever thanks to the --query
parameter. It replaces the other search parameters (-c, --time-range, --sat
) and allows for specifying full Data Store compliant open-search queries. The library DataStore object also exposes the new DataStore::opensearch(query)
method that will produce a SearchResults
object for the provided query.
Downloads have also been updated with progress bar and throughput displays (which can be disabled with the --no-progress-bars
) and new parameters that offer more control of where to download individual files. Use the --dirs
parameter to download each product into its own directory, or --onedir
to force everything to be together in the same output folder.
It is also now possible to get authentication tokens from the CLI directly with the new eumdac token
command. But before, configure the Consumer Key and Secret with the new eumdac set-credentials
command (the old eumdac --set-credentials
will still work).
Added support for local Data Tailor instances
This new EUMDAC version brings the capability of interacting with local Data Tailor instances. Avoid the limits on parallel jobs and workspace quota, these self-hosted instances are only limited by the capabilities of the system where they run.
Just setup your local Data Tailor following the installation instructions, add it to EUMDAC with the EUMDAC local-tailor command and include it in your download requests.
Find detailed information in the Data Tailor Standalone in EUMDAC section.
Improved error reporting
Though a lot of effort is put on building robust and reliable systems, errors can still occur. For when they do, the error reporting mechanism has been overhauled in the CLI.
When displaying error messages, the system will always include the HTTP status code, if available, and all the information gathered from the failing system. It will also try to be very clear about whether the issue was on the user side or the server side.
Learn how to extract all of this information from exceptions when using the EUMDAC library following the Exception Handling guide.
About EUMDAC
The EUMETSAT Data Access API Client (EUMDAC) simplifies the way users can discover, search, download, and customise data contained in the data store. Instead of writing lengthy code to download a particular product type for a specific time range, the user can do this with one or a few simple commands, invoking the client.
Two types of data access client are available:
- Python library — comes packed with methods and tips for using EUMETSAT's APIs and services. Just 'import eumdac' and begin coding.
- Command-line executable — allows users with little or no programming background to access the Data Store and Data Tailor from a command line, allowing the automation/scripting of jobs. Downloadable prebuild binaries ensure that users can use the client without any installation.
How to get it
The client is already available in the EUMETLab public repository, PyPI and conda-forge
Information on installation, usage and training material, including updated Jupyter Notebooks, can be found in the EUMDAC knowledge base.
For further information, contact our User Service Helpdesk.