Solved

DMT Job Assembly Update (Parent, Prior Peer, Next Peer, Child)

  • 4 February 2023
  • 1 reply
  • 93 views

Userlevel 4

Migrating via re-implementation to K22.1 via DMT

 

When we do the cut-over of the non-transactional Pilot to Live, then load the open jobs, I’m hitting a road block and can’t remember how to solve it.

 

What I expect:

Initially load JobAsmbl records with no parent, peer, or child fields included, and no Related Operation included.  Then load the JobOper records.  Then load-update JobAsmbl again, this time with only the required fields plus Parent, peers, Child, BOM sequence and level, and RelatedOperation.  The assembly records would then have their proper structure in the loaded job.

 

What’s actually happening:

 

On the Update run, DMT seems to load the update JobAsmbl records normally, no errors.  But the job assemblies are still flat in Job Tracker.  In one job, assembly 5 should be a subassembly of assembly 2 but it is still shown under parent zero.

 

What I’ve tried so far:

  • Load without Bom level and sequence fields
  • Load without RelatedOperation field
  • Run conversion workbench pointer fixes before DMT load
  • Disable my BPMs that “fix” the pointers, before the DMT load

Thanks for any input!

…...Monty.

icon

Best answer by mwilson 8 February 2023, 04:39

View original

1 reply

Userlevel 4

Solved our own problem.  The sequence is:

  • On the initial update load, include only required fields plus ParentAssemblySeq (ignore Parent; DMT has an issue using that field name because of .Net)
  • On this same load, add an Excel random field and sort by it, then remove it just before loading.  There are issues loading the assemblies if they’re loaded in order.
  • Re-random, re-sort and re-load as needed until the last of them work.
  • On the second update load, include only required fields plus RelatedOperation (no need to randomize this one).  To save time, filter out the zero related operation records.
  • Run conversions 750, 760, 770 in the Conversion Workbench.

Reply