We just started noticing this since the 05/09 update.  I believe that this is happening after running MRP, but not sure- we discovered the problem after running the Make and Buy reports.  We have had several lines of inventory where the O/H balance was zeroed out.  We found that on some of them the BKICLOC record was missing, while on others, the O/H field was 0.  On the ones where the BKICLOC record was missing, the MTICMSTR record was incorrect in the allocation field.  We used maintain database to correct the errors (in all cases, we had stock on hand and the transactions were correct).
Has anyone else seen this?

We are running 8/4/08

dbernhard-
 
    The book value is calculated seperately.  Therefore, if adjustments are made to inventory, the book value is not automatically updated.  UT-K-G updates the book value.
    As for SM-J-C, could you give some more information on how it is run?  Did you loose the inventory after a Master or Transaction Level Recon?  
    I run it here occasionally, and always in report only mode first.  I have found that running IN-L-S (Rebuild Stock Status) before SM-J-C seems to reduce the number of things that come up.
    Basically, there are several inventory master files, which can get out of synch (i.e. similar fields in two files can have different values).  SM-J-C allows you to correct this.

Lynn,
 
  I stand corrected on the book value.  I know that it used to not update if you made a change in IN-K, but it does now.

VMan -
 
Actually, if you change the Average Cost in IN-K then Book Value is not updated but you do get a warning.  It is NOT recommended to use IN-K to change the average cost of anything with on hand quantity.  better to use IN-C and make a negative adjustment to take them all out of stock at the incorrect cost, then a positive adjsutment to out them back in at the correct cost.  That will maintain Book Value as well as an audit trail of the change in cost.

Sorry to jump in the middle here, but I have a question...
 
You ran Master and Transaction level SM-J-C, and for the transactions did you specify that the on-hand should match the transactions or vice-versa?  
 
I made the mistake of running it in the mode that forces the on-hand amount to match the transactions and found that some on-hand amounts were changed when they should not have been. I seldom run transaction level reconciliation (for real) anymore. I run the master quarterly and run the transaction level as a report only to see if there are things that don't agree. If I feel it is needful, I might run the transaction level, usually only at fiscal year end, to make the detail balance in case the auditors look at it.

(I work with Debra)  To answer Colin's question, we run both a Master and Transaction Level, simultaneously, and use Method B.  We do this monthly.  We have found that our transactions are most always correct, and so we always adjust on-hand quantities to match the transactions.

Just had one burp on me today.  The BKICMSTR record for the part disappeared, but the BKICLOC record was still there.  Running SM-J-C master level would have deleted the BKICLOC record on a good part with stock O/H.  I re-entered the part in IN-B, and all is well again.  
I always run SM-J-C in report mode first so that I can look at what it finds. My experience has been that on the transaction level, you can have a mix between transactions being right and inventory count being right.  So I usually have to do some research after the report mode.