I think I may have identified a bug in the Database Compare tool.  Several (possibly all) of the integer columns in my model indicate that a datatype change has been detected upon every Database Compare.  See the attached screenshot. 

ModelRight won't actually allow me to either update the Database or Update the model (as you can tell from the screenshot, the Next button is grayed out).  Also, I can't identify any differences in the table definitions.

(My entire model is a product of the Reverse Engineering feature, and these columns have not been modified since the RE was performed.)

 

ModelRight, Version 3.0.0, Build 54, Meta Model Version 3.16

SQL Server 2005, Enterprise Edition (64-bit), Version 9.00.3042.00

 

Is this a bug or am I overlooking an option in the Database Compare feature?

 

I updated MR to Build 55, and I am still experiencing the same behavior.

Bump

Thanks for the bump.  We will investigate.  BTW, its easier for us to track bugs if you send them to [email protected].

---

Scott Reynolds
Senior Technical Support
ModelRight, Inc.

This is very odd.  Clearly they are the same, so why are they marked as different?  I cant reproduce either.  Could you try another/new RE and compare?  Could you try it with 32bit?  Any additional clues would be appreciated.

---

Scott Reynolds
Senior Technical Support
ModelRight, Inc.

I can certainly try another RE.  I will try to post something Monday morning.

I cannot, however, try it on a 32-bit server.  All of our SQL Server deployments, in DEV, TEST, and PROD, are 64-bit.  The workstations are all Win XP SP2 32-bit.

 I can provide a DDL if it will help.  I can also make myself available for a WebEx.

I have performed several additional REs, but I cannot reproduce the bug.  Each of the subsequent REs (each from the same server, same DB, and based on the same model) worked correctly.

The only thing I hadn't yet modified in the newly RE model is a trigger -- which did cause a problem in the old model.  When I modify a trigger (specifically, by assigning a schema, which was not assigned as a result of the RE), MR generates a DROP TRIGGER statement which fails because it does not include a schema (even though I checked "Schema Names").  This is an issue every time a trigger needs to be modified because every time the RE/Database Compare runs it fails to assign a schema.

This is a best guess as the cause of the issue because (I think) it appeared when we added additional triggers.