> -----Original Message-----
> From: Next generation catalogs for libraries
> [mailto:NGC4LIB_at_listserv.nd.edu] On Behalf Of Casey Durfee
> Sent: Thursday, July 19, 2007 3:25 PM
> To: NGC4LIB_at_listserv.nd.edu
> Subject: Re: [NGC4LIB] Open Source OPAC - VUFind Beta Released
>
> Excellent work. Very impressive.
Thanks Casey.
>
> It looks like your demo only has about 10,000 records or so. Have you
> tried running it against the Library of Congress MARC records
> (http://www.archive.org/details/marc_records_scriblio_net)? Doing so
> exposed several problems with both the indexing and display parts of
> the system I've been developing ( http://catalog.spl.org/catalog/ ), so
> you should definitely try that out if you haven't. Solr faceting
> performance scales pretty linearly with the number of records, so
> things that work fine on even 500,000 records can be unusably slow on 5
> million +.
We currently have about 550,000 records loaded. I was hoping to try it out with a larger dataset. I didn't know the LOC records were available. I guess I have been in a hole the past few months.
I have been playing around with your system a bit to see the performance of a large dataset. Doesn't seem too bad with the amount of records you have loaded.
Thanks!
Andrew
Received on Thu Jul 19 2007 - 13:37:38 EDT