DSXchange

Nandha,

The log file shows the correct cumulative CPU time, and you can also use that approach to get the values, if you were asking a question with your last post.

Anto, this is an OS level error, nothing to do with DataStage as you've already detected - if a ping won't work then DS won't either. Somewhere your network has either a misconfiguration or a firewall blocking your access to that ip address. You can trace part of the path by issuing a "tracert MIS_P...

Your P_AMT is coming from 2 different links, so DS has to choose one to drop. Often this is caused by enabling RCP.

Sure, the DSGet* functions will return values for PX jobs; in fact you need to be careful and parse the results correctly. In this case, teh DSJ.STAGECPU will return a comma-separated list of CPU times per processing node. But values are only returned for active stages, not for passive ones! Also, b...

I would recommend going to your backup and doing a clean restore if you can, especially since the cause hasn't been identified yet; it will guarantee a "good" version for your continued use. Just don't forget to make a backup copy of your current (broken) DS just to be 100% certain and safe.

That means that you probably haven't got some catastrophic failure in your projects. The Orchestrate checks aren't necessary yet, since you are having problems communicating with your project. Can you open up an administrator session to your project? Also, you might want to go into TCL as listed ear...

Bala, you will need to tell us what is or isn't happening when you try this before anyone can assist, also a "command" stage is a sequencer stage and a transform stage belongs in a normal job - so you do need to explain what you've done. If your source file is a sequential one, you can use the "filt...

Can you do the following commands from UNIX: cd `cat /.dshome` . ./dsenv bin/dssh LOGTO {YourProjectName} COUNT VOC QUIT This will attach to your DataStage engine home directory, set the environment and enter the the TCL prompt. Then it will do a DataStage level CD to your project and check to see i...

Hello Poorna, embedded in your error message is the message: ... ORA-00942: table or view does not exist ... which looks like the cause of your problem. If you are certain the table (fully qualified with schema) does exist the way you enterd it in the job I would recommend doing a deferred load. I d...

Pavan,

there are several possible causes for this, some only minor while others are more severe.

What has changed in your installation or happened to the system since you last logged in to the XXX_MI_PROD project?

Prashant, you have two problems and have asked two questions in your post - You have a job that is not working correctly and need to figure out why, and you have a process that you want to stop. Stopping your job should first be attempted by issuing a "stop" request from the director window. Sometim...

Holymac, there are two documented and supported method types of getting log information for DataStage jobs. The first is if you are "outside" of datastage and uses various call options on the dsjob command line to have the log information you want printed to standard output so that you can process i...

Vinod,

I would use the COMO command in TCL. This duplicates all screen IO to a file until turned off.



The file will be created in the project's &COMO& subdirectory as {TextFileName}.

Jim,

that was an interesting link! I've become so used to using MKS toolkit that I never even looked at the market to see if there was any competition. If you or anyone else here has used both could you make a comparison?

Hello Champa, the easiest way for you to get the log entries into a text file is to use the UNIX shell command dsjob . This command have several options that let you get just the information you want. If you call up the command without options you will get a help list, or you can read the documentat...