DSXchange

Some of us know exactly why you are doing it - to get a multi-row result set by creating the hashed file with X key fields and doing the lookup on less than X. Been there, done that, got the t-shirt. You need to understand how terribly slow that is without indexes and the restrictions that come with...

The version of database that I'm using is Oracle 9i. And I'm using the Oracle 10g client to connect to it and in DataStage, I'm using ORAOCI9 stage. I'll bet that's your problem right there. Why would you use a 10g client to connect to a 9i database? Think of it as using a 7.5 client to connect to ...

'My method' will not send out any more emails than any other method mentioned here. As noted, they would be coming from more locations in the job stream rather than the desired 'one central location' but you have control over what gets sent out, so it's not like you'll be drowning in emails or anyt...

Yes. 'Query' is not the same as 'lookup' and sql queries are horrible slow on a hashed file that was designed to do a singleton lookup. Search the forum for CREATE.INDEX and you'll find all kinds of discussions on the topic... this is just one example. Your biggest problem is going to be the fact th...

As mentioned, it's the long-ass option that is what was formerly known as 'Column Generated' in 6.x and 7.0, the 7.5.x equivalent is: 2. Generate SELECT clause from column list; enter other clauses We use that one exclusively, unless it can't be done via that mechanism at which point we fall back on...

Right... it automatically adds the TO_CHAR for selects and a TO_DATE for insert/update when the field is declared as a Timestamp in your job - and that *is* the proper type to use for an Oracle DATE field. I don't understand your problem. We do what you did in thousands of places and it is the way i...

Exactly. It ain't no tutorial.

The 'easy' way is to allow each sub-sequence to do its own mailing, then all you need is the 'Include job status' option checked.

What is the datatype of the Oracle fields?

Here is a link which explains how you can enter the contest.

Ok... if you are now saying that you only want to remove the 'X' from the field if it is the last character of the string, then you are right - you can't use a simple Convert function to do that. You'll need to do something like Ray posted, check the last character and if it is an 'X' then extract e...

It works if I treat the whole row as a single column and transfer the file. That's how I generally do it - and seems to me the best solution for preserving the struture of the file. The FTP stage is metadata driven and perfectly happy to massage data or change delimiters or do all kinds of other th...

Funny, but you're not the person with the problem. Still, what was your solution? We try to be all about sharing solutions, ya know.

chulett wrote:Don't know TeraData, but it seems the sticky point is this bit: "while in FastExport". Is that an optional something that can be turned off?

So... is that a 'no'?

Any idea why this error is coming. No not really, not with so little information SunShine. However, we can make some guesses. What kind of fields are in your source table(s)? Anything fun like a CLOB? If not a CLOB are there any NUMBER fields defined? Not NUMBER(xx) but just NUMBER... those are the...