DSXchange

The name is KIM. You must create the dictionary items. When you unzipped EtlStats then there should of been a folder called SqlScripts. There are a bunch of examples on how to create these dictionary items. There should be one called zDictCustom.sql in ColDicts folder. It contains this code. INSERT ...

I posted a new zip file for KgdGenHtml. It has source code on all but 2 routines. JobReport was not written by me. Tony Curcio wrote it and I do not have source. Maybe he will post it. Nice job Tony by the way. The other routine without code is KgdGenHtmlFromSqlRtn. I am not teaching you guys how to...

DSStopJob() is the API call to stop a job.

If you look at the generated SQL then you will see that it will only do an UPDATE ... WHERE KEY1 = ...

You should be safe.

I think Ray's point was to not redesign. It was to think in a new way. There is not a one to one comparison between how to do something in PX versus Server. It is a whole new set of stages and concepts. Somethings you cannot replace with a concept in PX. The whole purpose of the Transition Lab is to...

mailx -s "$SUBJECT" $USERS

Works fine for me.

XP Pro is not a supported platform.

EtlStats has a job named DSJobReportDb. It is in a folder called EtlStats\ETL_Row_Counts. It has a before job ExecDOS command: #dsjobcmd# #projName# #jobName# XML >KimD\#jobName#.xml Where in DOS dsjobcmd="..\..\Engine\bin\dsjob -report" If you have UNIX then this needs to change to a ExecSH command...

Not a good idea in my mind. When you upgrade then you need to upgrade all servers soon. You need to stop development until all servers or upgraded.

Q1 has been answered. Q2 use both todays date and the max date in your audit table. Q3 use max date + 1 till today -1 because today is probably not complete. Q4 You need a record in your audit table for each source. Insert this record only after today is done. A lot of people have a record like this...

The whole purpose of DataStage is to do this for you. You lose a lot of the beauty of DataStage when you go around it. The only reason to do this is to do a pure update. Please explain why you feel you need to go around DataStage.

It is so much easier to make all the columns varchar on a hashed file. Deal with timestamps only when writing to a real table.

Does your DataStage Administrator work?

You need to telnet into to your DataStage server. Ken posted the code to get to a TCL prompt. export DSHOME=`cat /.dshome` cd $DSHOME export PATH=$PATH:$DSHOME/bin . ./dsenv dssh The > is the TCL prompt. You need to Quit to get out of the DataStage engine. Next LOGTO MyProject I would do: PTERM CASE...

There are people who do this but if a job is running and you clear the log file it is just not smart. The phantom process is updating this table and rows disappear. If everyone is off the system and you do this then it is probably safe. The RT_LOGnnn files have surrogate keys. The next available key...