Re: A question about ParalellMultiSearcher and RMI

You may try to use the java parameter with -xmx to set up your memory. If you have 8G Ram in system , but java VM only caluse the smaller usage. And I try to merage the index from disk to RAM . When Index is only about 1G size , but into the RAM . I try to trace the memory usage , over 1G (about 1

Re: lucene 2.3.1 QueryTermExtractor error

That is the problem! -- an old highlight.jar Thank you, Erick! Erick Erickson wrote: Are you absolutely sure that you don't have some older jars hanging around in your environment? Especially when using Eclipse, the jar you compile with isn't necessarily the jar you run with, you have to find th

Re: How to Retrieve Found Term?

Edwin Lee skrev: > Hi all, > > i'm using Lucene 2.3.1. What i'm trying to do seems straightforward enough > (to me), but i just can't find the method to do so. > > Let's say i'm doing a PhraseQuery of the phrase "apples and oranges" with a > non-zero slop value, and it returns, e.g., 20 Hits. B

Re: lucene 2.3.1 QueryTermExtractor error

Are you absolutely sure that you don't have some older jars hanging around in your environment? Especially when using Eclipse, the jar you compile with isn't necessarily the jar you run with, you have to find the magic runtime settings And if you're not using eclipse, it could still be somethi

A question about ParalellMultiSearcher and RMI

I want to use RAMDirectory to raise the peformance of lucene. So I cut the index dir to 3 smaller index dirs(1G one index dir). Then I use RAMDirectory to load the 3 dirs, and rebind the 3 searcher by RMI.(My server has 8G mem) At last, I lookup the 3 remote searcher to new a ParalellMultiSearcher

How to Retrieve Found Term?

Hi all, i'm using Lucene 2.3.1. What i'm trying to do seems straightforward enough (to me), but i just can't find the method to do so. Let's say i'm doing a PhraseQuery of the phrase "apples and oranges" with a non-zero slop value, and it returns, e.g., 20 Hits. Because i'm using non-zero slo

Re: Occasional Hang in IndexWriter.close()

Hmm, that is odd. This is actually expected, if you are using ConcurrentMergeScheduler, because close() by default waits for all running merges to complete. But you seem to be using SerialMergeScheduler. Were there any exceptions hit before you call close()? Mike Stu Hood <[EMAIL PROTECTED]> wr