Large volume .trt file research help needed

[rkasnick]

Hi, I am developing a text document retrieval application which needs to search for and retrieve a specific document with sub-second response from a data file. A traditional SQL-type database is not acceptable due to licensing/support issues based on number of users this will roll out to. This is for a US Customs Import Rulings database. The rulings will number in the 100s of thousands and must have that sub-second retrieval I have heard that such an engine exists and uses DATA.INX and DATA.TRT files as the data repository, however I can not find anything on this site or the web et.al. which gives info on what this application development system is.

Does anyone have any information on how to create .trt data files? Thanks

 

[jrbarnett]

http://filext.com/file-extension/trt

http://filext.com/file-extension/inx

Have you looked into licensing SQL (or other DB servers) on a per processor basis, the number of concurrent clients won't matter on the licensing front.

John

 

[rkasnick]

John, I have looked at the link you sent, it is not the same, the file I am looking to create will be a text file, the link you sent was for a code base.

Thanks anyway for the response.

Licensed SQL will not work as all the users are not local and support for a formal SQL db would not be available at their locations.

 

[jrbarnett]

Web interface?

John

 

[Thargy]

I have to second John's idea, go on a per processor licence, and use oracle's built-in standard features for handling bulk text.

With that amount of data and the desired response time, a web interface seems ideal. Let the users get what they need and then they leave you in peace.

Not using a RDBMS seems franlky daft. They're designed and built for this sort of task, hence nifty things like partitioning to help deal with vast amounts of data.

Regards

T

 

[SQLSister]

I go with the others. Use SQL Server or Oracle or mySQL(or other large scale database) as backend with a processor liscense and a web interface for your users. If none of us know about the product you are talking about and it is isn't in common use, how hard is it going to be for you to get help when the inevitable issues of implementation arise? There are plenty of experts available for common databases and web programming languages. 100s of thousands of records is trivial for these databases (we have over 8 million records in our main table and ours would be considered a smallish to medium db) and you can't have more users than their current large users like Amazon.com or Microsoft.

"NOTHING is more important in a database than integrity." ESquared