Running quality checks using different hashstore instances

[doulikecookiedough]

Every repository that has upgraded to the latest Metacat version will be utilizing hashstore as the backend object store. This means that a dataset's objects and system metadata should no longer require API calls to the MN/CN to retrieve. However, the existing metadig.properties set-up to support hashstore has a hardcoded location.

To support multiple repositories, we need to consider how the metadig-engine obtains the location of a hashstore and its properties. One path could be the re-purposing of an existing redundant parameter (localFilePath) in the signature of the QueueEntry class when a quality check is being submitted to execute.

Investigate, discuss and then implement the changes required to allow a metadig-engine to support multiple repositories (ex. knb and adc) that have hashstore located in the same file system.

[jeanetteclark]

This is currently being implemented as a first phase (without rewriting config) to first try a hashstore, then try the API calls if the hashstore calls fail