Index Partitioning

[Shashi Kant]

Is there an "elegant" approach to partitioning a large Lucene index (~1TB)
into smaller sub-indexes other than the obvious method of re-indexing into
partitions?
Any ideas?

Thanks,
Shashi