DSXchange

Use the following condVal: if in.Id <> prevVal then in.Amount else condVal:"-":in.Amount prevVal: in.Id Intialize condVal to "". Specify four columns in your output. First three are straight pulls and the fourth column will have derivation condVal. Your data will look like this Id1, Type1,...

are you providing the fully qualified name. Try giving it with quotes. See if that works. The file is getting created, otherwise how is the data being loaded.

It means, to reset the job from director. Some more messages will pop up in your log file. Post them here.

Can you please post the complete error message. Doing a copy paste of the message usually helps.

And make it a practice to keep monitoring disk space allocated for your project directory. Also, avoid building files within project directory. Make sure logs are purged regularly.
My 2 centy.

Get the last event number using lognewest with the dsjob command. That will give you the number of entries present in your entire log file. If you want number of enteries for the latest run then you need to get lognewest before the job starts and after it ends. Subtract them and that will give you t...

In that case, true , it will not work. I was under the impression that each line is coming in as a seperate record, and hence my solution. Also, if the OP has input records coming in as one open, close tag in one line then it should be fine, else a more generic solution is needed. Agreed.

You sure Ken. The second Index will stop at the first occurance of "</". But that number minus the length of trailing tag will give you the length of the content between the tags. Try it.

narasimha wrote:

chulett wrote:In any case, switching to generated sql should solve the problem.

Craig, how did you infer this by looking into the error message any pointers...

Thats by burning midnight oil

Do you have the sample mass update scripts. Maybe you can put it up as an FAQ. This will help all of us a lot Ken.

PS: Loved the slurpee example

You sure you dont have your caps lock on ?

Do your unions and other transformations in DataStage. Since your source is a database, go for Ken's suggestions. Using mod, select data in chunks and pass it through your transformations, simultaneously, using multiple instance. This is for the "ET" part in ETL. For "L" , use a bulk loader for all ...

Ok then my code will do totally different. Use Handle_Null(). Refer here for more details.

I think you need to go over the ftp.pdf help document. It will tell you how to set it up and get it running.

NullToValue() will check if incoming is a null and then change it to the value you specify as second argument. Do this in the transformer