Indexing problems

[Sean]

Hello FBO group!

Firstly, let me complement you on an outstanding project! I've seen
the Laurentian version and it's really great. So much so, that we've
decided to try to implement it in place of our current OPAC Backup
Catalog (a MySQL, Perl, PHP beast of a thing).

I've read through the Project Documentation and the threads in this
group, but none seem to address the problems I'm running across.

I'll start with what I think I've done correctly:
1) Django appears to be up and running (on the included web server.
I'll switch it to Apache later). I'm getting a 404 error when I go to
localhost:8000
2) Solr appears to be up and running as well. I get the admin page at
localhost:8983/solr/admin.
3) I've checked out the latest SVN of FBO.

What appears not to be working:
Indexing. Solr contains no records after running the indexer on our
Marc21 extract. Instead, I get "Post failed for following record:"
followed by a line containing all the records in that batch. I've got
the output in a text file, and can post it in a follow-up or email it
(or if Google Groups can host files, I can do that, but I'm not used
to Groups yet).

Thank you all so much for your help,
Sean Moore
Tulane University

[Dan Scott]

Hi Sean:

I'd be interested in seeing the output; Google Groups does allow you to either attach files via email or to add a file to the group. I'd also be interested in knowing whether you're posting to Solr from the same machine on which Solr is running.

--
Dan Scott
Laurentian University

[Sean]

I've posted the output log to the group's files.

I'm running all the components (Solr, Django, FBO) on the same
machine. I don't know if it makes any difference, but I'm running it
on Windows Server 2003 (all my Linux machines were taken). I've also
commented out the memcached option in settings.py, as I didn't think
it was necessary for my test server.

Thanks!
Sean Moore

On Oct 27, 12:58 pm, "Dan Scott" <deni...@gmail.com> wrote:

[Dan Scott]

On 28/10/2007, Sean <TheDreadP...@gmail.com> wrote:

I've posted the output log to the group's files.

I'm running all the components (Solr, Django, FBO) on the same
machine.  I don't know if it makes any difference, but I'm running it
on Windows Server 2003 (all my Linux machines were taken).  I've also
commented out the memcached option in settings.py, as I didn't think
it was necessary for my test server.

Thanks!
Sean Moore

Hi Sean:

I think Casey was running the original code on Windows, so I doubt that's the source of your problem. I've eye-balled the first set of records that failed, and they appear kosher to me. Can you confirm that you're using the schema.xml file from indexer/config/schema.xml in your Solr configuration?

And yeah, the memcached option isn't necessary until you actually go into production (even then, it may not be necessary depending on your load). At the indexing stage, however, we're not even touching Django - it's purely Jython + marc4j + Solr - so we don't have to worry about that as being part of your woes, either :)

[Sean]

Aha!

I had transfered over the schema.xml and solrconf.xml files, but had
failed to restart solr (I must have started it before moving those
files for some reason). Indexing is working now.

My final question (hopefully!) is a Django one. It's running on the
included server, but only serves up a 404 page with this output:

"Using the URLconf defined in urls, Django tried these URL patterns,
in this order:

1. ^catalog/$
2. ^catalog/search/$
3. ^catalog/item/$
4. ^catalog/feed/atom/$
5. ^catalog/feed/rss/$

The current URL, , didn't match any of these."

Where should I point it to to serve up the html? Or did I miss
editing a settings file somewhere?

Thanks again!
Sean Moore

On Oct 28, 11:19 pm, "Dan Scott" <deni...@gmail.com> wrote:

[Dan Scott]

Hi Sean:

Glad to hear that you figured out the indexing problem!

Try http://localhost/catalog/ (the trailing slash is important); I'm guessing you pointed at http://localhost to start with. In production, you could set up your web server to automatically redirect any URLs that don't start with /catalog/ -> /catalog/ if you wanted to force all traffic to go to your catalog.

[Dan Scott]

A follow-up to my own message - I've updated the documentation wiki page at http://code.google.com/p/fac-back-opac/wiki/ProjectDocumentation to clarify the running the catalog / connecting to the catalog steps. I hope this helps :)