Unexpected reasons for failure

[vardhan354]

I'm doing an initial load Here is what I have done.
Source has 18.6 million records.

Method 1 :
Job1 extract to file (success)
Job2 Load from file (Failed)
lots of warnings and rejects
Error message :Inserted value too large for column, row rejected.
ORA-01722: invalid number.
I used the same meta data information for the sequential file column definition as that of the target table.
I have no clue for the error.

Method 2:
Direct load from database to database.
No warnings or rejects.
Only concern is snapshot too old may occur.



Any suggestions.....

[chulett]

Try using something other than a comma for the delimiter - like a tab (009) for instance.

[vardhan354]

Excuse my ignorance.How can I keep 'tab' as delimiter.Should I mention 009 in default padding section.

Thanks,

[ray.wurlod]

Surely ALL reasons for failure are unexpected. Do try to create more meaningful subjects!

What's in the file?

[kishore2ds]

Hi,
U can try this
In Job1:
Before loding into the staging target file,If u dont have then map one Transformer stage trim the columns as target mttadata lengths
If ColumnName1 contains length 38
Eg:Trim(ColumnName1[1,38])