DSXchange

Actually, the goal of the job control is to abort the job using DSLogFatal if the record count returned by the UNIX script is zero. How do I achieve this? By checking what record count the script returns and if it is zero calling DSLogFatal. Don't make it more complicated than it is. After calling ...

So, you are saying that we can use "pRCount" even without the last two lines in the job control shown. What I'm saying is that you've declared a Job Parameter called pRCount in your job control job. If you want to reference the value that parameter had when the job started, simply use the parameter...

-3 equates to DSJE_BADPARAM or "ParamName is not a parameter name in the job". This from the dsapi.h file in your $DSHOME/include directory. I don't see how this relates to your question. You are using DSGetParamInfo to see what the value of a parameter in your current job control job is - what doe...

ArndW wrote:You shouldn't be using SUBSTRINGS function, that is specifically for dynamic array strings.

AHA!

Stop using the fixed-width option. Or... don't forget that the Display attribute is what drives actual field sizes used. Post the details of your job design, the file itself and the actual errors you are getting if Ken's advise doesn't turn on the proverbial light bulb.

What data type did you use in the hashed file? I point out the 'ed' at the end before you-know-who comes along and does the same. If you used Date try sticking with a Varchar. I'd also suggest using the substring operator "[]" rather than the 'substrings' function, wish I could remember exactly why ...

Welcome. :D

Are you on a 7.5.x release of DataStage? If so, then you have the UserVariables Activity stage which can be used for this purpose.

:D Yes, most of us have seen them I'd wager. That 'wall plug' is the generic icon it displays for a "plug-in" stage when it has no clue what the real graphic should be. If you reboot or at least close out of the Designer session and come back in, is all well? Memory issues can cause it, I've seen ov...

So, a rolling two months? As in delete some of the data and copy some of the data? A detail not originally noted.

Then in that case... never mind.

I'd suggest you report it as a bug. Perhaps it's a known issue and there's a patch available for it? Stranger things have happened...

Assuming the formula is static and just that value can change, make it a job parameter. Or a single record hashed file pulled in at runtime in the Initial Value of a stage variable using something akin to 'UtilityHashLookup'. Or it can be echo'd in from a flat file. Or simply use a cached reference ...

The stage will always attempt to create a hashed file being written to if it doesn't exist. That option allows you to either override the default values used for the creation and/or delete the hashed file each run.

Rather than deleting everything from YYYY, copying everything from XXXX to YYYY, deleting everything from XXXX and then loading new data into XXXX - perhaps a table swap is in order? * Drop old YYYY * Rename XXXX to YYYY * Create new XXXX * Load XXXX The first two would need to be scripted but may b...

In 'automatic' mode, I assume? Not that I've found and it is definitely one of the downsides of that mode. Not only that but those kind of warnings don't seem to count against the limits you set at runtime. My jobs run with a warning max of 5 or less and I've seen an automatic bulk load happily log ...

Good to know! Thanks for posting that.