You are viewing a plain text version of this content. The canonical link for it is here.

Posted to dev@lucenenet.apache.org by juliya james <ju...@yahoo.co.in.INVALID> on 2018/04/05 04:09:33 UTC

Bigger index size produced with Lucene .Net 4.8.0 if compared with equivalent Java version

Hi,
The measurement of index sizes generated by Lucene .Net 4.8_beta005 shows that, it is bigger(~20-30%), compared to the index created by Lucene Java 4.8.0, as can be seen in the table below.
Are there any known concerns related to the index size with Lucene .Net 4.8.0?

                                  
| Index Size in MB withLucene .Net 4.8_beta005 | Index Size in MB withLucene Java 4.8.0 |
| 5.4 | 5 |
| 27.46 | 25 |
| 41.32 | 32 |
| 47.66 | 45 |
| 95.3 | 90 |
| 238.14 | 221 |
| 476.4 | 385 |



Note: -  IndexWriter configuration and the quantity of the data which was indexed(input) were the same for the measurements, so that the resultant index sizes can be compared. - While indexing, data was split to several Lucene documents and each document may have ~1MB of data. Most of the data was indexed with the field property   [Field.Store.NO, Field.Index.ANALYZED_NO_NORMS]
Thanks & Regards,Juliya