Additional text in the output of Execute Commant Stage

sumitgulati · Post by **sumitgulati** » Mon May 28, 2007 3:03 am

Hi,

I have a sequencer job that has three stages that are called in series
Stage 1: Job Activity Stage - It calls a multiple instance job (Job1)
Stage 2: Execute Command Stage - It executes a unix command.
Stage 3: Job Activity Stage - It calls another job (Job2) based on the output of unix command in previous stage.

I am facing strange problem. At times the output of stage 2 is appended by the text mentioned below
"[4444511] Done : DSD.RUN Job1.38 0/50/1/0/0"

Because of this the job called in Stage 3 does not get triggered as the output from the Stage2 does not match the required condition.

I am not sure why is this additional text coming intermittently.

Whenever I face this issue I simply run the job again without making any change to the code and it works fine.

Could you please through some light on this.

Thanks in advance for your help
-Sumit

JoshGeorge · Post by **JoshGeorge** » Mon May 28, 2007 3:50 am

Done : DSD.RUN indicates a run of a server job. Now why this is appending to your Execute Command Stage can be figured out only by knowing what exactly your are doing with Execute Command Stage.

sumitgulati · Post by **sumitgulati** » Mon May 28, 2007 4:35 am

JoshGeorge wrote:Done : DSD.RUN indicates a run of a server job. Now why this is appending to your Execute Command Stage can be figured out only by knowing what exactly your are doing with Execute Command Stage.

Thank you for your reply

In the Execute Command Stage I am running the following code:

grep 'TRAILER' #$FilePathDir#/#FileName# | wc -l

$FilePathDir is an environment variable for file path directory
FileName is a job parameter for File Name

This code will tell me if there is a record in the file that has text 'TRAILER'. If it is present then the output of the command would be the count of records that have text 'TRAILER'. If there are no such records the command should return zero.

The additional that I am talking about does not always get appended to the output of the command. It happens only sometimes.

sumitgulati · Post by **sumitgulati** » Tue Jun 05, 2007 3:32 am

Hi All, any suggestions on this issue.

Your help would greatly appreciated.

Regards,
-Sumit

chulett · Post by **chulett** » Tue Jun 05, 2007 4:37 am

Looks like output that would normally be in your &PH& directory. Best to call your Support provider and ask them, I would think.

sumitgulati · Post by **sumitgulati** » Tue Jun 05, 2007 5:28 am

Thank you Chulett.

ivannavi · Post by **ivannavi** » Tue Jun 05, 2007 5:55 am

Hey, I didn't know output of the command was available in Execute Command Stage! Only return code for the trigger.
Where can I find output?

sumitgulati · Post by **sumitgulati** » Tue Jun 05, 2007 6:12 am

ivannavi wrote:Hey, I didn't know output of the command was available in Execute Command Stage! Only return code for the trigger.
Where can I find output?

For Execute Command activity there is an activity variable "$CommandOutput" that stores the output of the command.

ivannavi · Post by **ivannavi** » Tue Jun 05, 2007 6:37 am

Thanks!
How stupid can I be? I know I saw it before (never used it) in the "Insert parameter value" list of any subsequent job activity!

ivannavi · Post by **ivannavi** » Tue Jun 05, 2007 6:37 am

Thanks!
How stupid can I be? I know I saw it before (never used it) in the "Insert parameter value" list of any subsequent job activity!

ivannavi · Post by **ivannavi** » Tue Jun 05, 2007 6:39 am

This is not my day. Now I post duplicate posts

ray.wurlod · Post by **ray.wurlod** » Wed Jun 06, 2007 12:39 am

I have seen something similar to this in the past, where DataStage (UniVerse) was running out of file units in the rotating file pool, and data destined for one file unit was misdirected to another. Were there very many files open at the time?

Does the problem occur reliably? If not, is it only at busy times and not at quiet times?