External string too long..Job Aborted

Post questions here relative to DataStage Enterprise/PX Edition for such areas as Parallel job design, Parallel datasets, BuildOps, Wrappers, etc.

Moderators: chulett, rschirm, roy

Post Reply
Developer9
Premium Member
Premium Member
Posts: 187
Joined: Thu Apr 14, 2011 5:10 pm

External string too long..Job Aborted

Post by Developer9 »

Hi,

While I am testing my jobs in latest version 11.3 I got below error ..In version 8.5 job is running fine (I am using same data file )

Code: Select all

seqfile >>>xfm >>>ds

Code: Select all

APT_CombinedOperatorController,xx: External string too long. Imported only 100 out of 125 external characters. Ignoring last 25 characters up to delimiter.
Please let me know if any one has seen this error message before

Thanks
qt_ky
Premium Member
Premium Member
Posts: 2895
Joined: Wed Aug 03, 2011 6:16 am
Location: USA

Post by qt_ky »

Yes, I have seen this many times when migrating to newer versions.

There is usually a metadata mismatch or some other detail in the job design, such as delimiter or other setting, that does not match the data.

In the older versions like 8.5 or 8.7, DataStage was more relaxed about enforcing metadata settings. In newer versions it is more strict.

Developer needs to pay more attention in newer versions and choose correct data types, lengths, etc. Hope that helps!
Choose a job you love, and you will never have to work a day in your life. - Confucius
Developer9
Premium Member
Premium Member
Posts: 187
Joined: Thu Apr 14, 2011 5:10 pm

Post by Developer9 »

Thank you for the suggestion ..When I ran the job in stand alone (not in through sequence job)job finished with too many warnings.

If there are no data issues ,How can we eliminate these warnings with out making the changes to the jobs ?As we need to make lot of changes the jobs (each job reading around 100-200 columns and all are defined as varchar)

Looking for the Options to have the Sequence job to finish with out abort .
qt_ky
Premium Member
Premium Member
Posts: 2895
Joined: Wed Aug 03, 2011 6:16 am
Location: USA

Post by qt_ky »

Warnings are there to alert you of potential data loss and the like. It is a good practice to go after the root cause and correct the jobs that process the data to prevent the warnings and ensure correct results through a wide range of testing.

To mask the problem, you could choose to change the sequence jobs to allow warnings or run with no warning limit or handle the warnings with a message handler. I think message handlers are usually a poor choice, to be reserved as a last resort.

Just to reiterate, these types problems have been there all along, but only the newer version is informing you. The problems are not with the newer version itself.
Choose a job you love, and you will never have to work a day in your life. - Confucius
Post Reply