On Aug 10, 2005, at 6:47 PM, Chris Fraschetti wrote:
maybe this is a stupid question, maybe not...
hits.id returns an int.. which would lead me to assume the obvious
limitations of the size of the index (size meaning number of docs) ...
assuming I reach this limit, can I expect lucene to throw some sort of
exception?
What is the best practice for this? watch my count and switch to a new
index when the time comes, then search across both indexes?
Who's got Lucene indexes nearing the maximum integer value? I'm
seriously curious. That's over 4 billion documents.
As for scaling - at that point I recommend checking out Nutch's
architecture, especially the new MapReduce branch. It's made for
this sort of big time scalability.
Erik
---------------------------------------------------------------------
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]
