DSXchange

Hello All,

I am getting the following error when using FastExport to extract data from a teradata table using MLaod stage.

**** 14:16:19 UTY8400 Error determining default character set name, code = 224

Does anybody have an idea what this error is about?

Thank you,
Satish.

I haven't seen that error before. However, have you tried using the Teradata Enterprise stage for retrieving data? That also uses FastExport for reads (FastLoad for writes). We have had much better luck using the Enterprise stage than any of the other API stages.

Brad.

bcarlson wrote:I haven't seen that error before. However, have you tried using the Teradata Enterprise stage for retrieving data? That also uses FastExport for reads (FastLoad for writes). We have had much better luck using the Enterprise stage than any of the other API stages.

Brad.

Thanks Brad. Now i am using the Teradata EE, it works fine. But we are extracting about 10 million records and i am afraid about the performance using Teradata EE.
If you have any heavy volume extraction using Teradata EE version please share with me, that would greatly help.

Thanks,
Satish.

Teradata EE is likely to give you the best performance (throughput) of all the Teradata stage types. Are you taking proper advantage of parallelism?

Hi Ray,

Would you mind elaborating further a bit on the parallelism parameter settings ?

Hi,

When using TeraData Enterprise stage be sure to use "DBOptions". Like this :
user=myuser,password=*******,SessionsPerPlayer=2

you can add as well, the "RequestedSessions" option.

There is a relation between the "SessionsPerPlayer", the number of nodes and the resulting TeraData Sessions generated. In the above example ( SessionsPerPlayer=2) you will have 16 TeraData sessions for a 8 DataStage nodes.
In our environment ( a 20Tb TeraData database ) we try to limit up to 15 Teradata sessions for each application.
Try to balance performance and TeraData sessions by tuning the SessionsPerPlayer.

Hope this help.

Great answer from Daniel - not much I can add to that.

smuppidi wrote: Thanks Brad. Now i am using the Teradata EE, it works fine. But we are extracting about 10 million records and i am afraid about the performance using Teradata EE.
If you have any heavy volume extraction using Teradata EE version please share with me, that would greatly help.

Thanks,
Satish.

Satish, FastExport is the Teradata tool of choice when retrieving data - low or high volume, and that is the underlying tool used by the Teradata EE stage.

Brad.

Thanks all, great information. I could get that working fine using the Teradata EE.