I don't think I can do a table-to-table import because I need the 100 different serial numbers on the 150 components (well, 15,000 components I suppose). But by using a temp table to put just the 150 components in from the larger Master Data table, the pipeline sped up a lot.
Since the pipeline may get triggered more than once during the long duration of a run, I made sure the temp table had the PCBA ID in it as well as the time the record that triggered the whole thing was updated. The pipeline still does a search on the temp table and it looks for both the PCBA ID
and the timestamp. That way, even if there are multiple triggering events during the first event's run, the multiple "instances" of the pipeline will get the right data from the temp table based on the timestamp. Also, the pipeline deletes the data from the temp table using the PCBA ID
and timestamp.
This is the biggest set of data manipulation I've done in Quickbase and it definitely made me think about multiple users and/or pipeline instances, giving a sense of "ready to do the next thing" to the user, and managing large sets of data!
------------------------------
Edward Hefter
www.Sutubra.com------------------------------
