DSXchange

Try http://www.dsxchange.com/viewtopic.php?t=89641.

Do a search. There has been several discussions using stage variables.

I was told that uv executes as root and sets the ulimit then runs uvsh if you are on UNIX. So unless you are accessing large files then you can execute uvsh directly. The uvsh command is like sqlplus for Oracle. It is an easy way to execute a SQL script and redirect the output to a file. This would ...

I don't type it in. I drag it to the target and then drag it back. I am not sure when you could drag from right to left from the target derivation to a lookup key field but it saves time. It is more accurate too. I am uncomfortable with it being red anyway.

How "open" will it be? Does anyone know?

Michael

Even Ray uses this feature. Don't you Ray?

I think you could modify the routine in http://www.dsxchange.com/viewtopic.php?t=89419 post and change a parameter value in the current job.

You could of corrupted an index. DS.TOOLS can rebuild the index. Do a search. This is also one of the FAQ.

I think what they want is to archive logs to some kind of flat file. There have been several solutions posted even recently. Do a search for RT_LOG and archive.

You need to find out the difference between thier date in internal format and DataStage's.

NewDate = oconv(TheDate - 104314 + iconv("2004-11-09 ", "D-YMD[4,2,2]"), "D-YMD[4,2,2]")

Something like this should work.

dsjob -report Project Job XML Will output job start and end times and row counts. You may want to download EtlStats.zip from my tips page or ADN. I process the output of this command and route it to several tables to store current and historical row counts. As far as executing the next job then all ...

You can also use convert() and compare it to the original field to look for a specific character like "&".

if InLink.ColumnName <> convert("&","", InLink.ColumnName) then @true else @false

This is a server routine. When you save metadata in a link then it gets saved to DS_METADATA. If you know how many records are in a table and you haved saved or imported the metadata then you can use this to estimate disk space. This would work on sequential and hash files as well. I have never want...

You need 2 lookups. Have 2 outputs with constraints for each situation. Another way maybe using stage variables with if STATUS='A' set and second set of assignments if that is not true and the second situation 'STATUS' is either 'A' or 'B' or 'C' is true.

You may need to kill -9 PID. Normally you can just kill PID but in your case force the kill command.