DSXchange

And there's the fact that they are stored in the export in (what I seem to remember and assume is) MIME format so not really representative of their actual storage.

And the subject of their reply (which you have to start a reply to to see) was "thanks a lot" so perhaps it worked?

Pretty sure that was 8.5 but someone else would have to confirm. Or deny.

Sure but that wasn't the topic of discussion. Besides, by pretty darn small I basically meant zero.

See if this works in the Connector stage.

While I'm sure people can puzzle it out, please start by detailing what you pass in for "arg1" and what the error is that you get so we don't have to. I'd also be curious why you set a FLAG to zero and pass it out as the ANS without ever changing it to a non-zero value when there's a probl...

Care to share the actual error?

I've been bitten in the past by that approach as it banks on the surrogate being always incrementing and that the highest number is latest record and there can be circumstances that cause that to not be true. It basically assigns some level of "intelligence" to the surrogate whereas I pref...

sathyak098 wrote:all the records which are inserted on this particular load

How will you know which records those are? Are you 'marking' them in some fashion?

8) On the subject of "at most 2^32-1 values", I just wanted to chime in with the thought that (in my mind at least) it's not a concern for the application we're discussing. If you were trying to generate "unique" values like I've seen some people do in a misguided attempt to esta...

srds2 wrote:I tried this conversion also String to TimeStamp(Column,'%m/%d/%yyyy %h:%n:%s.6 %aa'). This gave me an error saying invalid timestamp format.

So not so invalid after all?

I wouldn't assume anything (especially anything 'significant') but rather do a quick test to see what kind of speed you get from the CRC32 function and one comparison v. comparing all data columns. We're doing this using MD5 on sometimes a couple hundred columns (in another tool) mostly because it i...

By chance could you post an "ascii art" picture of the stages in your job? So we are certain how the data is flowing from source to target. Wrap it in

If you are on 8.7 then these statistics are already being captured for you. To get "incoming" for a stage, check the "outgoing" from the immediately previous stage.

Filter on CATEGORY.