You need to tell us a bit more about how Lucene will be used. You can, for example, keep each DB in a separate Lucene index and use MultiSearcher or you can put all the data into one Lucene index. Both of those have their advantages and disadvantages so until when we know more about your system, I don't think we can help much.
-- George Aroush -----Original Message----- From: rzr rzr [mailto:[EMAIL PROTECTED] Sent: Thursday, March 22, 2007 1:23 PM To: [email protected] Subject: Indexing single-file databases Hi, I've got a couple of questions. I'm posting them in two seperate mails to keep the subject headers relevant. Given: Relatively large databases that contain many records. Each database consists of a large file which may contain as much as 2M records. Each record consists of several fields, though exaclty what fields and how many of them may vary. The records are seperated by a NewLine/Cr/CrLf combination. Question: I need to be able to search and rapidly load ALL positive records. How would you suggest I procede? (A link to a solve or an article would also be welcome in lieu of an answer.)
