From: Wayne G. <ws...@wm...> - 2008-03-13 20:48:22
|
I'll add a script soon, but basically you need to do this to use your Solr instance: java -Xms256m -Xmx256m -Dsolr.data.dir=/path/to/solr/data -jar vufind...jar The Java code actually skips passing the code through the server interface and instead writes the index directly. Wayne Andrew Nagy wrote: > Oh ... with the new import system - since we are still in test mode - the data doesn't get directly imported into Solr. It creates a new lucene index. You should now see a new solr directory in the import directory. You just need to delete the old index and move the new index in place. > > Let me know if you need any help with this. > > Andrew > >> -----Original Message----- >> From: Jeffrey Barnett [mailto:jef...@ya...] >> Sent: Thursday, March 13, 2008 3:54 PM >> To: Andrew Nagy >> Subject: Re: [VuFind-General] Index errors in 0.8 import >> >> Well the real problem was that the catalog.mrc contained marcxml (from >> 0.7 instance) not true marc. When I re ran with raw marc, the records >> were imported into solr OK, but the wrong version of solr. I'm still >> running 0.7 on 8080 and I started a second jetty/solr instance on 8090 >> to contain 0.8 content. I set and exported %JETTY_PORT=8090, but >> apparently there are some other pointers to reset? >> >> Andrew Nagy wrote: >>> Jeffrey - I experience this problem too with a dump from our voyager >> catalog that included holding records. The importer is designed to >> deal with bib records only. I just ran the importer 2 days ago with >> our marc dump and it worked just fine. It just reported these errors >> for the holdings records. Are you including the holdings records in >> your export? Are records being loaded into Solr? >>> Andrew >>> >>>> -----Original Message----- >>>> From: vuf...@li... [mailto:vufind- >>>> gen...@li...] On Behalf Of Jeffrey Barnett >>>> Sent: Thursday, March 13, 2008 2:45 PM >>>> To: vuf...@li... >>>> Subject: [VuFind-General] Index errors in 0.8 import >>>> >>>> I just built an 0.8 version of vufind and the java importer. >>>> vufind.sh started solr successfully on a new port but when I run >>>> >>>> $ java -jar vufind-0.8-dev.2008.03.12.15.10.17.jar >>>> >>>> All I get is a console full of: >>>> ... >>>> Error indexing >>>> org.marc4j.MarcException: unable to parse record length >>>> at >>>> org.marc4j.MarcStreamReader.parseLeader(MarcStreamReader.java:317) >>>> at >> org.marc4j.MarcStreamReader.next(MarcStreamReader.java:138) >>>> at org.vufind.marc.MarcImporter.main(MarcImporter.java:190) >>>> Caused by: java.lang.NumberFormatException: For input string: >> "dataf" >>>> at >>>> >> java.lang.NumberFormatException.forInputString(NumberFormatException. >>>> ja >>>> va:48) >>>> at java.lang.Integer.parseInt(Integer.java:447) >>>> at java.lang.Integer.parseInt(Integer.java:497) >>>> at >>>> org.marc4j.MarcStreamReader.parseLeader(MarcStreamReader.java:315) >>>> ... 2 more >>>> ... >>>> for various input strings. Our marc records are utf-8 encoded, >> could >>>> that be a problem? Is there a setting or config I am overlooking? > > ------------------------------------------------------------------------- > This SF.net email is sponsored by: Microsoft > Defy all challenges. Microsoft(R) Visual Studio 2008. > http://clk.atdmt.com/MRT/go/vse0120000070mrt/direct/01/ > _______________________________________________ > VuFind-General mailing list > VuF...@li... > https://lists.sourceforge.net/lists/listinfo/vufind-general > |