DSXchange

Has this job been running in the past. A quick read of the query results indicates that such problems arise if something goes wrong with the executables, which can be fixed after a re-compile. The production support is right, they will not compile it, you need to recompile the job and export it again.

Makes sense. Thanks for the clarification Ray.

Check point restartablility will fire from the point where it failed last. If you want to skip that and go to the next one then you will have to code that part yourself. If there are no dependencies between the jobs, then run them in parallel, this way if job three fails, atleast job 5 and job 6 wil...

You can do this using stage variables. Sort your incoming data and then do the comparison between present key and first. Refer to this post.

Do an exact search on 'Unable to open the operating system file'. One thing is strange, the routine 'DataTypePicS9' is a datastage routine. It should be catalogued by DSX prefix rather than DSU.

thurmy34 wrote:Program "JOB.1118171632.DT.1414830161.TRANS1": Line 152, Unable to load file "DSU.DataTypePicS9".

Open command is equivalent to 'Before Sql' in Server and Close command is equivalent to 'After Sql'.

Did you see any error message/warning message in the logs. Did the job even fire ? If you do a search on error code '81015', you will see that it usually means there is a time out from the server. Need more details.

If you could do this in the database stage itself then you dont need to worry about lookup stage and the extra date condition. If you have to implement it in datastage then do a lookup on the second table, get all the results and constraint it in the transformer using the date range. Something like ...

If you could re-do the Basic Routine in C that would be great. This way you can call that routine in a 'normal' transformer. This is offcourse if the data volume is expected to be large.

Yes, but as Craig said

chulett wrote:Right, something very much like that - but without the hard-coded timestamps.

So how come the DELETE command was working without a semi-colon

Sort it at once, in the begining on all the grouping keys, in that order. Then you can use that dataset and pass it through different aggregations for different groupings.

I know, just playing with ya !!!

Only specify the sort as asc for the first column and sort it by the first column. That row is out of sequence, needs proper sorting.

Really, where....