Umbrella Issue: Service Monitoring, Compliance, and Reporting in CKAN

[mwengren]

This is meant as an umbrella issue to plan if and how to build a monitoring capability into IOOS Catalog/CKAN. We have several existing issues that are related:

• IOOS catalog data services health monitor catalog#60: IOOS catalog data services health monitor
• Create a service compliance report in Catalog #204: Create a service compliance report in Catalog
• ERDDAP dataset attributes to read for CKAN filtering, API, and Schema.org output #208: ERDDAP dataset attributes to read for CKAN filtering, API, and Schema.org output

What these all have in common is some functionality that reads dataset service endpoints within the IOOS Catalog inventory, and perform some sort of synchronous (check compliance - #204, read attributes - #208) or asynchronous (ioos/catalog#60) action on them.

Maybe the synchronous/asynchronous distinction divides them too much, but if not, can we think of some ways - perhaps even using external cloud services like Lambda - to read the Catalog database periodically for appropriate service types (specifically ERDDAP tabledap and gridddap and THREDDS OPeNDAP) and run simple checks against them:

• see if they're up once per day
• test compliance using Compliance Checker

and record the results somewhere they could be displayed in the Catalog UI or made available via other means.

For the attribute query (#208), it makes sense to do this as part of a harvest cycle, since we would want to update a change to key attribute immediately (e.g. ioos_ingest=True|False).

Perhaps all of this functionality needs to exist in Python alongside the CKAN codebase, but let's consider other possibilities that leverage cloud services as well.