KVRF Workflow for Automated Data Acquisition (KWADA)
This is just a proposal. The assumption is that this processing is triggered by saving data to a shared directory that looks like:
<sharedDir>/metadata/xx1.csv, xx2.csv, ...
<shareDir>/images/xx1.
I Metadata
1. Save Metadata to Shared Directory ?
2. Alert Auto Metadata loader process (backup existing database) ?
3. Load Metadata ?
4. Run Indexer.
Steps 1 to 3 of I must occur before 3 of II.
II Image Workflow.
1. Scan Images ?
2. Save Images to Shared Directory ?
3. Generate Derivatives ?
4. Create ImageId / MetadataId map ?
5. Load Map ?
6. Goto I 4.
This assumes that data is loaded into the 'Live System'. If some sort of confirmation that the data is 'good' is needed, we need to stage the data, confirm integrity, and load production system from staging area. In that case, there is a third flow.
III Staging to Production flow
1. Confirm data integrity ?
2. Export data from staging ?
3. Backup Production System ?
4. Load staging data to production system