Tek-Tips is the largest IT community on the Internet today!

Members share and learn making Tek-Tips Forums the best source of peer-reviewed technical information on the Internet!

  • Congratulations Chris Miller on being selected by the Tek-Tips community for having the most helpful posts in the forums last week. Way to Go!

(TR1901) A record has been rejected. An arithmetic overflow has occurr 1

Status
Not open for further replies.

ms061780

Technical User
Feb 19, 2007
4
EU
Can anybody help me?
When the creation of data cube is finished, a failure is appearing (see a part of the log file) and the created cube, of course, shows no values.
What does the failure mean and what can I do?
I have tried it out with a subset of data and it works fine.
Only when I use the complete data the failure occures.
The cube is about 160MB (I do create bigger and more complicated cubes as well without any problem).

Timing, TOTAL TIME (CREATE CUBE),00:40:57
(TR1901) A record has been rejected. An arithmetic overflow has occurred. (Data) in d:\cubes\imq\tat_customer.mdc. [->OK]

I'm using PowerPlay Transformer(7.3.1221.0 ).
 
Looks like one of your fields that Transformer has been instructed to treat as a numeric field contains a non-numeric entry. Your subset is not including it, so looks 'clean'.
If you can't correct the item on the database and be sure it wouldn't re-occur at source, then you'll need to error-trap the data as it comes into Transformer. Some DBs provide an 'isnumeric' function that could be used here.

soi la, soi carré
 
Hello drlex,
Thank you for reply!
But when I share the data in two subsets I can construct the cube on both of the parts. All data are included eigther in the first or in the second subset.
If a value in a numeric field would be not numeric, I could not create a cube with parts of the data.
 
In that case, a quick check on the Cognos knowledgebase suggests you might also:

1. Check for the Storage type in "Measure" properties. It should be able to accommodate the measure data, if unsure, set to 64 bit floating point.

2. Check if there is a case where a single category has more than 65536 children in the source data. If this is the case, modify your source so that it breaks down the children into several (artificial) subcategories (for example based on the first letter of the category name), so that there is no category having more than 65536 children.

doc 102656.10 Feb 2007, so a recent addition.

(There's also a reference to the problem occuring when
using DB2 on AIX, corrected with a later client version - 1026417.1).

soi la, soi carré
 
Status
Not open for further replies.

Part and Inventory Search

Sponsor

Back
Top