This is something you would have to configure in postgreSQL. I believe the default is to use btree, but you should be able to change it to hash. Of course you'd have to rebuild your indices after the change.
Sue Walker-Thornton Software Developer/Database Administrator NASA Langley Research Center|LITES Contract (757) 224-4074 From: Franck Borel [mailto:[email protected]] Sent: Monday, December 13, 2010 3:59 AM To: [email protected] Subject: Re: [Dspace-tech] Index row size exception when starting index-init Hi Keith, thanks for your answer. Franck - Is your repository dspace.cfg configured to create, from the description.abstract field - a Lucene search index, configured with the search.index parameters, used for the search function in DSpace OR - a PostgreSQL index, configured with webui.browse.index, used for the browse function in DSpace. I was wrong: Not description.abstract.field make update-handle-prefix/index-init failed, but the title field. I have imported the items via batch file and some titles are extremly long. Is there any reason why DSpace uses B-tree instead of Hash for the PostgreSQL index? If it's configured for both, do you really have a need to browse (not search) by the description.abstract field? Does the handle update work when you disable the browse configuration for description.abstract? You could run the SQL that updates the handles manually, but I suspect you would still have problems later when your index-update process runs. Yes. The simplest solution I found was to shorten the title via script. I have updated the metadatavalue of the author field. Now it is not empty anymore. The error: "2010-12-10 11:53:40,074 ERROR org.dspace.browse.IndexBrowse @ Null metadata value for item 1119, field: dc.contributor.author " doesn't appear anymore. The second error is a result of that the B-tree index in PostgreSQL has a maximum size of 2713 bytes. I am working on a project that includes the need to store a fair amount of raw text in the dc.description.abstract field. It seems that I am forced to outsource the abstract into an text file and then upload this file into dspace... :( Has anyone an idea how to resolve this issue? -- Franck Have a nice day! -- Franck ------------------------------------------------------------------------------ _______________________________________________ DSpace-tech mailing list [email protected]<mailto:[email protected]> https://lists.sourceforge.net/lists/listinfo/dspace-tech -- Franck Borel Albert-Ludwigs-Universität Freiburg Universitätsbibliothek UB 2 Abteilung EDV Rempartstrasse 10-16 79098 Freiburg Telefon: +49 761 203-3946 Telefax: + 49 761 203-3889 E-Mail: [email protected]<mailto:[email protected]> Internet: www.ub,uni-freiburg.de<http://www.ub,uni-freiburg.de> <franck_borel.vcf>------------------------------------------------------------------------------ _______________________________________________ DSpace-tech mailing list [email protected]<mailto:[email protected]> https://lists.sourceforge.net/lists/listinfo/dspace-tech ------------------------------------------------------------------------------ Oracle to DB2 Conversion Guide: Learn learn about native support for PL/SQL, new data types, scalar functions, improved concurrency, built-in packages, OCI, SQL*Plus, data movement tools, best practices and more. http://p.sf.net/sfu/oracle-sfdev2dev _______________________________________________ DSpace-tech mailing list [email protected]<mailto:[email protected]> https://lists.sourceforge.net/lists/listinfo/dspace-tech -- Franck
------------------------------------------------------------------------------ Oracle to DB2 Conversion Guide: Learn learn about native support for PL/SQL, new data types, scalar functions, improved concurrency, built-in packages, OCI, SQL*Plus, data movement tools, best practices and more. http://p.sf.net/sfu/oracle-sfdev2dev
_______________________________________________ DSpace-tech mailing list [email protected] https://lists.sourceforge.net/lists/listinfo/dspace-tech
