I have built a SOLR Index on Windows 7 Enterprise, 64 Bit. I copy the index to Centos release 6.2, 32 Bit OS.
The index is readable and the application is able to load data from the index on Linux. But there are a few fields on which FQ Queries dont work on Linux , but same FQ Query work on windows. I have a situation where in i have to prepare index on windows and port it on Linux. I need the index to be portable. The only thing which is not working is the FQ Queries. Inside the BlockTreeTermsReader seekExact API, I have enabled debugging and system out statements scanToTermLeaf: block fp=1705107 prefix=0 nextEnt=0 (of 167) target=1RD0JIHMr9aw4RPPuS0DVzB2tKf38FfjKaEg7HsYDd7EtAOpE9FYvvj5ryB7679r4KNnlIazevPo h7qabtLhXw== [31 52 44 30 4a 49 48 4d 72 39 61 77 34 52 50 50 75 53 30 44 56 7a 42 32 74 4b 66 33 38 46 66 6a 4b 61 45 67 37 48 73 59 44 64 37 45 74 41 4f 70 45 39 46 59 76 76 6a 35 72 79 42 37 36 37 39 72 34 4b 4e 6e 6c 49 61 7a 65 76 50 6f d a 68 37 71 61 62 74 4c 68 58 77 3d 3d] term= [] This is a Term Query, and target bytes to match As per the algorithm it runs through the term and tries to match , now the 6th term is a exact match, but there is a problem of few bytescycle: term 6 (of 167) suffix=1RD0JIHMr9aw4RPPuS0DVzB2tKf38FfjKaEg7HsYDd7EtAOpE9FYvvj5ryB7679r4KNnlIazevPo h7qabtLhXw== [31 52 44 30 4a 49 48 4d 72 39 61 77 34 52 50 50 75 53 30 44 56 7a 42 32 74 4b 66 33 38 46 66 6a 4b 61 45 67 37 48 73 59 44 64 37 45 74 41 4f 70 45 39 46 59 76 76 6a 35 72 79 42 37 36 37 39 72 34 4b 4e 6e 6c 49 61 7a 65 76 50 6f a 68 37 71 61 62 74 4c 68 58 77 3d 3d] Prefix:=0 Suffix:=89 target.offset:=0 target.length :=90 targetLimit :=89 from the first comment 50 6f d a 68 37 from the second comment 50 6f a 68 37. The test scenario is the index is built on linux and i am testing the index through solr api on windows machine. -- View this message in context: http://lucene.472066.n3.nabble.com/Portability-of-Solr-index-tp4061783.html Sent from the Solr - User mailing list archive at Nabble.com.