difference between fastload using a script and using teradata studio express

Teradata Studio
Enthusiast

difference between fastload using a script and using teradata studio express

Hi,

 

Right click on the table name and choose "import data", I can upload a csv file to teradata.

I got some questions:

 

1. How does studio upload this file to TD? is this fastload?

 

2. I tried to use load a file (less than 1000 lines) using the fast load script and I can see the using the script is much faster then using the studio.

Can you please explain why?

The data only has several hundreds lines so the session number of the fast load might not matter.

 

Thanks in advance,

Jing

 

 


Accepted Solutions
Teradata Employee

Re: difference between fastload using a script and using teradata studio express

I don't have numbers comparing JDBC FastLoad and JDBC Batch Insert. But we have an outstanding issue to gather such information so hopefully we will have something to post soon. As I mentioned, for small row count,  less than 100,000 rows, JDBC batch insert is used. When we do use the JDBC FASTLOAD option, we don't change the number of sessions and rely on the JDBC Driver to manage the FASTLOAD connection. Please refer to the JDBC Driver documentation for more information.

1 ACCEPTED SOLUTION
4 REPLIES
Teradata Employee

Re: difference between fastload using a script and using teradata studio express

In Teradata Studio Express, if the file has more than 100,000 rows, it will use the JDBC Fastload option, else it will use JDBC batch insert.

Enthusiast

Re: difference between fastload using a script and using teradata studio express

Thanks fgrimmer,

 

Is there a benchmark of the performance between batch fast load and batch inserts?

For a small row count, will these two have almost same performance?

Also what is the default parameters like session number for studie express fast load?

 

 

Teradata Employee

Re: difference between fastload using a script and using teradata studio express

I don't have numbers comparing JDBC FastLoad and JDBC Batch Insert. But we have an outstanding issue to gather such information so hopefully we will have something to post soon. As I mentioned, for small row count,  less than 100,000 rows, JDBC batch insert is used. When we do use the JDBC FASTLOAD option, we don't change the number of sessions and rely on the JDBC Driver to manage the FASTLOAD connection. Please refer to the JDBC Driver documentation for more information.

Enthusiast

Re: difference between fastload using a script and using teradata studio express

Hi fgrimmer,

 

I found a very good artical explaining the difference between JDBC insert and JDBC fast load.

http://developer.teradata.com/connectivity/articles/speed-up-your-jdbcodbc-applications

 

My question is:

Since fast load needs a big JVM heap size, I suppose this JVM is running in my local laptop. if I set the -Xmx with a low value, it will cause the uploading to fail. Am I right?

Also I am wondering if the low value will impact the performance of uploading?