Question

I’ve an SSIS dtsx package which is used to load data from a remote MAS db server using a DSN based connection. We load data from many tables into their replica tables in SQL-Server. Everything was working fine until we made some changes to a table in MAS. The dtsx has been failing with the following error:

Error: 0xC02090F8 at Data Flow Task, Import Data, DataReader Source
[28866]: The value was too large to fit in the output column
“UDF_TREAD_DEPTH” (29160).

Actually I believe it might be related to a single table field “UDF_TREAD_DEPTH” which is a decimal field. This field is shown in the DataReader source as "numeric [DT_NUMERIC]" with Length:0, Precision:4 & Scale:2.
In past we had simple data in format xx.xx. And now I see after the issue that we have data like xx.xx, xxx, .. however, still the data type didn’t change after I refreshed the Data Reader source.

I believe the “Precision shud be updated to 5” for the data we have
based on this description.

I’m unable to change the data type as visible in the attached screen (Data Source Output column.png). When I debug this dtsx package, it errs while loading the Data Reader Source. If I’m nailing it right – how can I fix it. If there’re any other possibilities then kindly let me know.

Answer 1

I got the data to be updated with the xxx.xx mask so 100 became 100.00. And this helped the DataReader in SSIS infer the type correctly.

In addition to it I also found another easy way of doing so which didn’t require support of any cast / convert function –

UDF_TREAD_DEPTH * 1.00 as UDF_TREAD_DEPTH

This also allowed the DataReader to infer the type (i.e. precision & scale) correctly.