I need a new database engine, can you help?

Question

Greetings,
I have a fairly unique application that I need to find a new database for. It currently uses an embedded database but we now want to go client server with it.
Background:
The application downloads a large amount of data, millions of records, from a database that runs on an IBM iSeries system. The purpose of the application is to allow the main database to be used at remote locations when a connection to the iSeries is not possible. At a later time the app uploads changed records. We want to go client-server so that the download only has to go to one PC and then other PC's can share the database across a network that our clients will set up at the remote location.
The trick is in the download performance. When we download from the iSeries we get a binary file that we can read records out of using Java's IO funcationality, one record at a time. We want to avoid having to do a SQL INSERT statement for each of the millions of records we need to place into the database because this tends to be slow. Ideally I'd like to be able to use Java IO to write the database files.
Do any of you know of a client-server database that allows this to happen?

Answer 1

The trick is in the download performance. When we download from the iSeries
we get a binary file that we can read records out of using Java's IO
funcationality, one record at a time. We want to avoid having to do a SQL
INSERT statement for each of the millions of records we need to place into the
database because this tends to be slow. Ideally I'd like to be able to use Java
IO to write the database files.Most of database(Mysql, Oracle, and so on) will provide dump/load utilization to do that task, and you need to use Java IO to format your binary file to the predefined binary format(maybe csv), then use Runtime.exec() method or batch to import all rows.
One special client-server database is HXTT Text(http://www.hxtt.com/text.html), which can use directly your binary file as database. If you needn't to merge daily millions of records into the same table, you can try it with its support from http://www.hxtt.com/support.jsp?product=text .