[OODT-725] CAS-PGE should not pass global metadata keys from Workflow to CAS-Crawler from reserved key set

• Type: Bug
• Status: Resolved
• Priority: Major
• Resolution: Fixed
• Fix Version/s: 0.7
• Component/s: crawler, pge wrapper framework, workflow manager

Environment

XDATA translation

• Skill Level: Don't Know (Unsure) - The default level
• Assignee: Chris A. Mattmann
• Reporter: Chris A. Mattmann
• Created: 02/Aug/14 18:44
• Updated: 04/Aug/14 23:05
• Resolved: 04/Aug/14 23:05

Description

Currently CAS-PGE adds all Workflow global metadata to CAS-Crawler before invoking it when it runs its ingestCrawler. This can cause a big issue on WorkflowNotifications from CAS Crawler e.g., by its TriggerPostIngestWorkflow or UpdateWorkflowStatusToIngest actions b/c they pass along to the workflow dynamic metadata the "Filename" keyword. This is great for staging the input for the next set of workflow tasks, but it messes up ingestion since CAS-PGE passes this information to CAS-Crawler before ingesting the outputs of the CAS-PGE task. Having the upstream Filename met field screws up the Crawler since it thinks that the filename for the ingest product is actually the preceding one instead of the one generated and being crawled.

CAS-PGE should NOT pass dyn workflow metadata to CAS-Crawler's global metadata on reserved keywords like Filename, FileLocation, ProductName and ProductType.

Issue Links

• is depended upon by: OODT-728 Figure out why certain metadata fields get set twice during pipeline processing
• requires: OODT-726 Create MetFilter Task Example