DSXchange

kduke

I think I would change the metadata on that column to a varchar. I would also have the job delete and recreate the hash file so you get a new set of dictionary items.

kduke

Dan

You are welcome.

kduke

I gave you a routine to find one file. You should be able to find the next 39. We are not here to do your job. Just point you in the right direction.

kduke

Do a search. This code has been posted before. StartedDateTime = DSGetJobInfo(RunHandle, DSJ.JOBSTARTTIMESTAMP) FinishedDateTime = DSGetJobInfo(RunHandle, DSJ.JOBLASTTIMESTAMP) StageNames = DSGetJobInfo(RunHandle, DSJ.STAGELIST) LinkNames = DSGetStageInfo(RunHandle, Stage...

kduke

At TCL run: select EVAL DS_JOBOBJECTS."@RECORD<6>" AS FILE_NAME FMT '35L', DS_JOBS.NAME AS JOB_NAME FMT '35L', DS_JOBOBJECTS.NAME AS LINK_NAME FMT '35L' from DS_JOBOBJECTS, DS_JOBS where DS_JOBOBJECTS.OLETYPE in ('CHashedOutput','CHashedInput') and DS_JOBOBJECTS.OBJIDNO = DS_JOBS.J...

kduke

I a routine you can open the directory like a hash file.

Code: Select all

Ans = @FALSE
DirName = Arg1
FileName = Arg2
openpath DirName to MyDir then
   read Rec from MyDir, FileName then
      Ans = @TRUE
   end
end

kduke

Make this a sticky in the Server forum too.

kduke

I think is a great idea.

kduke

Why not pipe this to grep. A filter can be several commands or a shell script with this in it. grep can look for more than one thing by putting them in a file. You could easily use awk or sed or even Perl.

I am not sure about filter in PX. I was thinking about Server jobs.

kduke

Ray should know. It is called KgdCreateDdlMetadata. It is based on a routine Ray posted.

Here is how to use it. http://www.dsxchange.com/viewtopic.php? ... dlMetadata

kduke

You are welcome.

kduke

Lance chill. Nature of the beast.

kduke

If you do a search for upgrade or migration then you should find some topics. I did some more searches and found these.

http://www.dsxchange.com/viewtopic.php? ... hlight=3.6
http://www.dsxchange.com/viewtopic.php? ... =Version+4
http://www.dsxchange.com/viewtopic.php? ... =Version+4

kduke

Aggregators are very slow and eat up a lot of resources. I would break them up into one Aggr per job. Also do them in a temp table in the source or target would also help.

kduke

There are posts by either Ray or Ken which explain the highlights of the changes.

Search found 4605 matches