Engine Warning: One Or More Values Were Out Of Range And Truncated For Node. WriteForecast threw an exception caught (Doc ID 734728.1)

Last updated on MARCH 28, 2016

Applies to:

Oracle Demantra Demand Management - Version 7.1.1 to 7.2.0.1 [Release 7 to 7.2]
Information in this document applies to any platform.

Goal

Find the following warnings in the the EngineManager log

ERROR WriteForecast threw an exception caught for node [1,332272,225206]

Before going any further in this note please run the following sql in the Demantra schema/database:

  -- select max(datet) from inputs;

If the value is 2010 then follow the actions in <Note 551814.1> and after that is completed then run the engine again to see if this issue re-occurs.  If it does then continue troubleshooting with this Note

Which leads us to the children Engine2k log where we find:

12:29:55:140 Machine localhost receiving branchID 1
12:34:30:671 <SAV> 12:34:30:671 WARNING one or more values were out of range and truncated
for node [1,332272,225206]
12:35:13:625 <SAV> 12:35:13:625 WARNING one or more values were out of range and truncated
for node [1,372197,245303]
12:37:50:296 <SAV> 12:37:50:296 WARNING one or more values were out of range and truncated
for node [1,337993,225206]
12:37:50:296 <SAV> 12:37:50:296 WARNING one or more values were out of range and truncated
for node [1,338304,225206]
12:37:50:296 <SAV> 12:37:50:296 WARNING one or more values were out of range and truncated
for node [1,338207,225206]
12:40:42:281 <SAV> 12:40:42:281 WARNING one or more values were out of range and truncated
for node [1,351136,229782]
12:46:10:671 Engines have finished executing.

The symptom here is that the Engine generated forecasts for some of the  nodes had been set to 10,000,000. There are historical records in sales_data.actual_quantity well over 10,000,000 (they are being stored correctly). In this case the client has very high volumes in their business so they need the forecast series to accommodate the larger values.  To see the solution for this type of client then please go to the next section.

However, for clients that don't deal with very large volumes like the values mentioned above the cause of the issue is most likely a Demand database stream issue.

We need to find what kind of Demand is causing the engine to write either these very large positive or very small negative numbers. To do that you need to go the root expression for 'Demand' that the engine uses and see if you have very large or very small numbers in for certain items indicating that your demand stream into Demantra might need to be corrected.
   
The engine looks at the Quantity_Form expression which is found by going to Business Modeler
--> Parameters --> System Parameters --> Engine --> Data Manipulation
   
Once you have the expression that is at your site (since quantity form can be different expressions at different clients)
   
(eg. nvl(pseudo_sale,actual_quantity)*(1 + nvl(demand_fact,0)) )
   
then run the following sql:

Once you determine the column(s) that is(are) causing the issue then correct your process (ex. user overrides or data feeds) to make sure these very large or very small (ex. volumes that are in decimals or negative)

Note that if your issue is negative numbers then please review <Note 464124.1> as to some thoughts on why negative values might be are coming into Demantra.

Solution

Sign In with your My Oracle Support account

Don't have a My Oracle Support account? Click to get started

My Oracle Support provides customers with access to over a
Million Knowledge Articles and hundreds of Community platforms