DSXchange

There are too many ways to do this to answer, at least given the information in the request. If your column length is always 1 and starts with "A" you could index with a simple SEQ(In.Col)-64 to get a number. If you column is of variable length then this method won't work, but if all columns are of ...

So now you know the problem is the READ. What read method are you using? What happens if you go to single node? Is your DB/2 database partitioned?

Assign JobHandle = DSJ.ME and then write a loop that uses a call to DSGetJobInfo(JobHandle,DSJ.JOBCONTROLLER) to get the parent job, then open up that job and issue the call again until you have found the ultimate parent process and then use another call to DSGetJobInfo to retrieve the start time. E...

Stiwari - what is your source and what is your target? How many nodes in your configuration? When you say that no records are being read are you sure that this is not because of an inability to write - meaning change your output to be a /dev/null sequential file or a PEEK stage.

The shared container is like a normal high-level language INCLUDE or INSERT statement and gets compiled as part of the job. DataStage does put the IDENT{nn} stages automatically between two passive stages (in this case the TeraData and Hashed file stage. Because of the way the the shared container g...

You can do what you want - but if, in your designer, you had pulled the connecting arrow in the wrong direction it reverse the direction of data flow to what you expect. What happens if you substitute a sequential file output stage for the dataset in a copy of your job? Do you get a similar error?

OK. LEN('x':CHAR(000):CHAR(000)) is 3 but you are getting 0. Either DataStage has a bug in your case and is adding two characters to the record or you might be looking at the wrong data. Could you check to make 100% certain? Is this a server job (if it is a PX job then a possible cause is the row ou...

The Naked Fish brings back lots of memories. Some of them are even good!

...>##E TFDR 000001 07:02:49(006) </home/mfarid/ds/Output.ds> Data set initialization for "/home/mfarid/ds/Output.ds": This input-file data set does not exist: /home/mfarid/ds/Output.ds... DataStage thinks that this is an input stage, i.e. a source, and not a target. Are you sure of your design and...

There can be a hidden file in the root directory called ".dshome" which contains the fully qualified path to the DS server directory. This ends with "/DSEngine". This file is not present when multiple DS engines are installed on the same machine. Since the file is missing and you only have one engin...

psv, you can see the additional text by becoming a premium member. The costs really aren't that high considering the time savings (perhaps not for this issue, but for future ones). Did anything go wrong with the install? I think it is worth trying to get someone with root password to try connecting,...

Can you explain your lognewest error, it sounds like a jobrun syntax error. You can use the dsjob -logdetail to get details for a given event id and can loop through all events in a loop looking for a job start. Event IDs are numeric and older entries have lower numbers.

Is this a new installation or did it, at some time, work? Have you tried connecting from some other client PC to see if the problem is the same? If you have root access, try logging in as root (to see if it might be access related).

Very strange behaviour. If you output In.Colum[1,1] does it still add the two CHAR(000) places?

Even shorted derivation: