RE: [greenstone-devel] Re: PagedImgPlug
| From | Emanuel Dejanu |
| Date | Thu, 30 Dec 2004 11:34:54 +0200 |
| Subject | RE: [greenstone-devel] Re: PagedImgPlug |
| In-Reply-To | (35813418-58B0-11D9-BEF6-000D93B1CDEE-bigpond-com) |
|
Try not to use hash for generating OIDs. From may
experience this takes a lot of time.
try
import.pl -OIDtype incremental
This are all the values that -OIDtype can get (I use
dirname):
hash: Hashes the contents of the file. Document
identifier will be the same every time the collection is imported. incremental: A simple document count that is significantly faster than "hash". It is not guaranteed to always assign the same identifier to a given document though and does not allow further documents to be added to existing xml archives. assigned: Uses 'D' plus the value of dc.Identifier as the document identifier. dc.Identifiers should be unique. If no dc.Identifier is assigned to the document, a hash id will be used instead. dirname: Uses 'J' plus the parent directory name as the identifier. This relies on there being only one document per directory, and all directory names being unique. E.g. import/b13as/h15ef/page.html will get an identifier of Jh15ef. Best regards,
Emanuel Dejanu
From: greenstone-devel-bounces@list.scms.waikato.ac.nz [mailto:greenstone-devel-bounces@list.scms.waikato.ac.nz] On Behalf Of stephen.degabrielle@bigpond.com Sent: Tuesday, December 28, 2004 11:10 AM To: greenstone-devel@list.scms.waikato.ac.nz Subject: [greenstone-devel] Re: PagedImgPlug If it is disk accesses slowing you down and assuming you have heaps of ram - you could try doing the imports from a ram disk. - just make small batches and do 2 or three at a time. Try import by hand to see if that works better - or gives you a better idea of whats going wrong. Remember to set the -verbosity to 3 for the most feedback. s. On 28 Dec 2004, at 5:56 PM, ~{R&~}~{g2>}~} wrote: Hello all, | |