DSXchange

Like I said, the forum software removes 'extra' spaces from what it thinks is normal text, so you need to wrap your example in Code tags to preserve the whitespace. Do the same for any actual code you need to post to preserve any formatting / indenting as well. You need to know what the total size o...

Sure, I understand. Just wanted the point clarified as I have zero experience with the CFF stage so am following along so I can learn things about it... vicariously.

Just out of curiosity, there's nothing inherent in a Server job that makes it "slow" just because it is a Server job. For example, I've used one to parse 10 million Apache log records in less than four minutes. Is it specifically the CFF stage that is the bottleneck? How long did it actual...

Back in the day when I was doing RTI stuff, we went the "heartbeat" route as well.

There's a user guide pdf near the bottom of this page. Sounds like you have full control over how many messages and what types you can unload.

Don't see any padding in your example, without 'code' tags all excess whitespace is removed by the forum software. What have you tried? And you need a space between Col1/Col2 but not between Col2/Col3? That would be... odd. Assuming no. What you seem to have is a fixed-width file with no spaces betw...

Your current way is the correct way and the only thing 'clumsy' about it is using a file for intermediate storage. Make your first job a Server job and put the output of the proc into 'user status' which is one of the public variables you have access to. Then you can link directly from the first job...

Yes.

Interesting. For grins, what was the source system of the file and what UNIX does your DataStage server run?

"control characters"? Can you post some examples? First suspect would be either your source data or how it is being read.

kumar66 wrote:But DB2 is storing the date formt as 13-Jan-2010 10:12:34 AM.

No, it is not. A timestamp is a timestamp is a timestamp, it is your query tool that is enforcing the external format of that timestamp when you select it.

If what you need is a full outer join, then you'll need to figure out how to do it, simple or not. As noted, it certainly looks like that's all you need to do - all except for the 'sid' part. Where does that come from, another table or do you need to generate it during the run? If you're uncertain a...

One of the tables will be your input stream, the other a reference lookup. If you need multiple rows from the lookup, you'll need to use ODBC or UV rather than the Hashed File stage for the reference lookup.

If you're trying to find the word "Header" in the more output, you need to enclose it in double-quotes in the actual command you run.

You could always take up little burden yourself, it's a good learning opportunity. Plenty of references for UNIX commands out there, including how to use 'find' for files of a certain age and then feed the results to other commands (like 'rm') via the '-exec' option.