DSXchange

If you make sure that you have closed all sessions with your user and the explorer shows active sessions with your user id then you have identified the culprit and can stop that dead session in the explorer.

If you make sure that you have closed all sessions with your user and the explorer shows active sessions with your user id then you have identified the culprit and can stop that dead session in teh explorer.

wellison - I didn't ask that the job gets changed permanently, just make a copy and change the datatype to see if this is the root cause of your problems.

Since timestamp are not string types and don't get converted, I would guess that this error occurs in your source stage reading StartTime from the file and then trying to convert it to a timestamp.

try a test with the password being a string and not an encrypted parameter, does the error go away or stay?

If the data is sorted and partitioned on the column with the duplicates, then use two stage variables, "IsDup" and "LastValue" derived via "IF In.ColumnName=LastValue THEN 1 ELSE 0" and "In.ColumnName" respectively, the constraint would be "IsDup=0" to only pass on non-duplicate rows.

Sort your data on Column A, then use stage variables and constraints in a transform stage along the following format: NewCol = IF(In.ColumnA=LastColumnA) THEN 1 ELSE 0 OutColumnA = IF NewCol=1 THEN LastColumnA ELSE '' LastColumnA = In.ColumnA OutList = IF NewCol=1 THEN ListString ELSE '' Lis...

With which error message does the job fail?

are the "<" and ">" characters coming as part of the encrypted values? Then you could just quote the string when passing it to the shell script. You will not be able to un-encrypt those values into cleartext, though.

Make a copy of the job, put in a transform stage whose output does not have RCP and remove a number of columns. Does the same error occur at the same location? Or dump rows 169670 onwards into a peek stage or sequential file and see if you can spot a possible cause in the data. These are just two ap...

Your input parameter date is a string. In a transform stage, preferably in the stage variables, convert the string to a date. Then apply the DateFromDaysSince() function to the date and use that in the transform. If you use the "initial value" for the stage variables then they will only be computed ...

I don't see a quoted string problem, but it seems that you use "column_name1" and "col1" - are these legal column names?

DataStage has mail sending capability within the email activity stage. DataStage will use the UNIX sendmail interface, so the question of attachments is just a matter of finding out how to do it with sendmail and then implementing that in a DataStage call. If you add some specifics to your question ...

Go into the IS Admin explorer tool (http://YourServer:9080) and remove the dead session, then try again.

I'm not at a workstation where I can test this now, but if you cannot use the user-defined variables directly, then you declare them as part of your job parameters, call them with a value of "$PROJDEF" and use them in the transform stage.