Updating the metadata server

Running a metadata server is the most complex part of the safedata system. It involves setting up and configuring a website and database and so requires further skills and resources. There are real advantages in running a metadata server: it makes your data much easier to find and use through the safedata R package.

However, you may choose to get started with using the safedata_validator system alone to ensure data standards and publish the datasets. You can always set up a server and populate it further down the line. Ideally, you should hang on to the dataset metadata files that are generated by safedata_validate and the Zenodo metadata files from publishing datasets. Both of these files can be recreated though, by re-running the validation on published datasets or using the safedata_zenodo get_deposit subcommand to re-download the Zenodo metadata.

Posting data for new published datasets

If you are maintaining a safedata metadata server then, after a dataset has been published, you should post the metadata to the server to make it discoverable using the safedata package. You will need to have configured safedata_validator to include the required metadata server information but then the safedata_metadata command line usage is:

# Post the zenodo and dataset metadata to the metadata server
safedata_server post_metadata zenodo_1143714.json Example.json

Updating the server resources

The metadata server is also used to distribute some of the validation resources used both safedata_validator and safedata, including the gazetteer files and the Project ID database. When you update these resources for validation locally, those new data also need to be posted to the metadata server. There are no command line arguments for this command - it simply updates the server to the existing configuration file.

safedata_server update_resources