tOracleOutputBulk -tOracleOutputBulkExec

[Anonymous]

Hi,

I have a file csv that I want to load in a Oracle Database by using tOracleOutputBulkExec because it is very big. Can you sho me a solution please?? 

This is my job

[TRF · Accepted solution]

Hi,

I suppose the output flow shcema for the tMap component is aligned on the tOracleOutputBulkExec.

So, what's the problem? You just have to link the tMap to the tOracleOutputBulkExec using the flow definde in tMap.

Or do you have a specific problem not exposed in your initial question?

[Anonymous]

Yes, i've link them but talend send me a message and then i don't understand what is name file in tOracle.

[TRF]

As for any tXxxxxOutputBulkExec component, the records are written into a file by the component itself using the format expected by the database or application you want to write into. So you don't have to know this format, just concentrate on what you want to transfer, not how to do that.

So, the filed "Filename", is the name of this file, not the name of the input file (and sometimes, there is not inputfile, for example if the source is an other database).

What is the message you received?

Is it a compilation error or a runtime error?

[Anonymous]

I don't understand again. 

My job is:

and my file csv is in this position:

My tMap is:

And i lin it with tOracleOutputBulkExec, but i don't understaind again the filename:

and talend send me this error message:

I have install Oracle in the Virtual Machine and now i'm working on the localhost. The component tOracleOutput is ok but is is very slow so i want use the bulkOracle.

Thanks again

[TRF]

When you try to use tOracleOutputBulk operations, you need to install 1st sqlldr which is the dedicated Oracle tool to load external data into databases. This should solve the error you get when running the job (cannot run program "sqlldr").

Regarding the filename in tOracleOutputBulkExec, just use the filename you want (not the name of an existing file).

This file will be used by sqlldr when called by the job at runtime.

[Anonymous]

If i run the job from VM but talend send me a message: unable to open file .dat . I have insert a filename without extencion.

[TRF]

As I remember, the default extension for sqlldr is ".dat", so name the file according to this rule.

[Anonymous]

Hi,

I'm sorry if i answer you so late. Thanks again for your help!

Now I have another problem with the bulk load . In particolar I have a problem with the timestamp format. When I load the timestamp with the tOracleOutputBulkExec, the timestampo contain hour, minutes and second set on 00:00:00.

[TRF]

I suggest you to open a new topic for this new challenge.
Also, indicate the expected format for this field.