Segmentation Violation?

[tomcmacdonald]

ERROR:  An exception has been encountered.
Please contact technical support and provide them with the following traceback information:
 
The SAS task name is [DATASTEP]
Segmentation Violation
 
Traceback of the Exception:
 
 
ERROR:  An exception has been encountered.
Please contact technical support and provide them with the following traceback information:
 
The SAS task name is [DATASTEP]
Segmentation Violation

I'm getting this message while trying to parse a ~2GB CSV file using UNIX SAS 9.4.

[ballardw]

With code there is a chance we could identify some problem that might cause that message. Without not much to guess from.

By any chance did you try to use Proc Import and see if that would work? If so you might want to start with the data step code generated by import and modify as needed. If you go this route I suggest setting a large value for guessingrows to get the best fit for your data.

The size of the CSV file probably doesn't have an impact though there is a possibility that content might.

[BennettR_]

This suggestion sounds plausible, except that the

     ERROR:  An exception has been encountered.

     Please contact technical support and provide them with the following traceback information:

     The SAS task name is [EXPORT  ]

     Segmentation Violation
comes BEFORE the generated code can ever appear in the log.  🤐

[Kurt_Bremser]

In your opening post, the TASKNAME was DATASTEP, but now it is EXPORT? 

Maybe you have a larger problem.

Please post the whole log of the failing step.

[jimbarbour]

@tomcmacdonald,

I've seen that error before on a number of occasions.  I think @ballardw's suggestion is a good one:  Start with a Proc Import and modify the generated code as necessary.   Usually there's something subtle going on, and it's hard to pin down when you get that error.  It could be some odd combination of characters and delimiters or the way an Informat is handling something or ...  who knows?  It's usually a lot easier to start fresh than it is to track down exactly what is wrong with one's data step.

Jim