DSXchange

I need to read an input sequential file(.csv) after first 5 records.How do I skip those first 5 rows while reading the file.

You have to read them, that's the nature of sequential media. You can skip them in the job using a constraint along the lines of:

... or you can skip them in Unix using the 'tail' command as a filter on your sequential file stage:
HTH,
J.

Make sure you run in sequential mode for craigs solution. If thats not what you want then provide the following filter command to restrict the first 5 rows at the sequential file stage itself.

If the first five rows don't match the metadata of the remaining rows, you can get away with doing nothing, just allow the import to fail for these rows (though it will log a warning message). You can capture these import failures via a Reject link on the Sequential File stage, thus allowing your to reconcile the row counts.