Hi Hossein: On Mon, 05 Jan 2009, Hossein Rafighi wrote: > I used the BibHarvest Admin Interface. Since my indico site uses ssl > and my original setup for base url in invenio was non-ssl I thought > that may have something to do with it, but adding a https to base url > made no difference. Please see the attached file.
The OAI harvester in CDS Invenio 0.99.1 supports harvesting from HTTP sites only, see <https://savannah.cern.ch/task/?8004>. Your first OAI source (the non-HTTPS one) as defined in the screenshot is using the h-u (=harvest-upload) mode, not the h-c-u (=harvest-convert-upload) mode. This is wrong and accounts for the empty record problem in your Invenio site. You have to convert harvested OAI XML files into MARCXML before uploading them, so that the h-c-u mode is needed here, similarly as you have defined it for your second OAI source (that is HTTPS though, which was not supported in v0.99.1, as mentioned above). (BTW, note that you have also an option to make more XML input file manipulations via a custom record filtering script, if you want, by using the h-c-f-u harvesting mode, but this should not be really needed. The h-c-u mode should be enough for your needs, provided that your Indico instance generates suitable MARC output. We use the h-c-u mode here at CERN too.) Please see also /opt/cds-invenio/var/log/bibsched_task_1234.{log,err} for more details about your harvesting errors. Here, 1234 is the task ID of your oaiharvest process. > Happy new year. Thanks and all the best to you and everyone :) Best regards -- Tibor Simko ** CERN Document Server ** <http://cds.cern.ch/>
