Hello Everyone, We have a legacy system which uses lucene 2.4.1. We have ported a small hack to lucene source code back then, so that the underlying lucene segment merger code wouldn't reuse deleted docids. This helped us use lucene docids as persistent dbids as well. But we want to upgrade lucene to 3.6, but it is near impossible to "hack" lucene now to get the same behavior.
I checked out NoMergePolicy, and it seemed to help achieve similar behavior of not letting lucene reuse deleted docids. But I guess this would increase the number of segments in the index. Any idea how many segments we are talking about over here? Also, can we configure lucene to tell how many documents to keep in a given segment. Each lucene index in this system can have utmost 1M documents in them. Is there an alternative that I can consider? Thanks, Snehal -- View this message in context: http://lucene.472066.n3.nabble.com/Usage-of-NoMergePolicy-and-its-potential-implications-tp3996630.html Sent from the Lucene - Java Users mailing list archive at Nabble.com. --------------------------------------------------------------------- To unsubscribe, e-mail: java-user-unsubscr...@lucene.apache.org For additional commands, e-mail: java-user-h...@lucene.apache.org