1/29/18: {Monica} sent info to Steve

we can generate multiple free SSL certificates through Let's Encrypt so that every IEDA repository would have the necessary certificate per sub-domain, so this component specifically would not be a cost-barrier to setting up multiple nodes. The certificates are 90 days at a time but we just set a crontab for running the renewal check periodically and forget it.

Software Stack¶

GMN + Adapter.
IEDA is a candidate for new functionality which will deploy multiple nodes from the same web app, using a different baseURL for each node. The adapter in this situation will need to be able to pick up metadata from different locations and assign the correct formatId in the process of creating in GMN.

Replication¶

None. IEDA picks up metadata from source repos that use landing pages for data access
points. Direct links to data bytes are not contained within, so replication is not an option.

Harvest endpoints:¶

• US Antarctic Program Data Center (USAP): http://get.iedadata.org/metadata/iso/usap/
• Marine-Geo Digital Library (MGDL): http://get.iedadata.org/metadata/iso/mgdl/
• EarthChem Library: http://get.iedadata.org/metadata/iso/ecl/

IEDA Work Plan¶

4/25/18- Dave check-in with Steve reminder of items he needs to send (see IEDA work plan)

5/3/18 {reply from Steve}
Took some reengineering to get all the partner systems working.
Here's a test example dataset landing page with the sdo markup, including the link to the iso record:
http://get.iedadata.org/doi/test/isosdo/100700.

Sitemap updates are next. The webmaster at IEDA will have to deal with the robots.txt files. WE haven't registered these formally with google yet, just informal with Chris Sater at Google.

Steve communicated some confusion regarding deployment. Misunderstood that we would still be deploying a GMN installation on web app with schema.org as intermediate step that lies between our developing the capability of the CN to communicate directly through the embedded metadata without the need for the slender node web app in between. A few major points from Dave's response:

1. We would prefer not to support a custom mechanism for advertising availability of the ISO metadata. That approach is a dead end that requires custom implementation and ongoing support from DataONE by way of operating a custom Member Node implementation.

2. We could support access of the content via OAI-PMH, however this too suffers from ongoing operational costs for DataONE since the intermediary Member Node will be required indefinitely.

3. The schema.org approach is preferred since even though there is an initial cost to implementation and deployment of an intermediary Member Node, there is a future path to completely removing the intermediate Member Node and synchronizing the content directly from the IEDA resources. The long term cost of this approach will therefore decrease significantly once we complete support for schema.org synchronization.

• sitemaps for the IEDA partners set up, with a sitemap index at http://get.iedadata.org/sitemaps/
• The links get html views of the ISO metadata records with the schema.org JSON-LD scripts in the page head sections.
• Haven’t put the link in robots.txt, this is still prototyping.
• A testbed is set up to generate the SDO markup, including link to the ISO xml record. Here are some example URLs (look at the page source to see the JSON LD in the header):

http://get.iedadata.org/doi/test/isosdo/100700
http://get.iedadata.org/doi/test/isosdo/302963
http://get.iedadata.org/doi/test/isosdo/500107
http://get.iedadata.org/doi/test/isosdo/609594

the 6 digit number is the token part of the DOI for registered resources.