splitting lucene index into two halves

Question

what is the best way to split an existing Lucene index into two halves i.e. each split should contain half of the total number of documents in the original index

Answer 1

A fairly robust mechanism is to use a checksum of the document, modulo the number of indexes, to decide which index it will go into.

Answer 2

The easiest way to split an existing index (without reindexing all the documents) is to:

1. Make another copy of the existing index (i.e. cp -r myindex mycopy)
2. Open the first index, and delete half the documents (range 0 to maxDoc / 2)
3. Open the second index, and delete the other half (range maxDoc / 2 to maxDoc)
4. Optimize both indices

This is probably not the most efficient way, but it requires very little coding to do.