I also am wondering if it doesn't like it on a local server. Do you have a public facing THREDDS catalog link that one of us could look at?
If so, I would be willing to try to harvest your dataset using GeoNetwork 2.10. There are definitely some issues with the way Thredds and netCDF data is pulled into GN (does not pull in all the metadata* - it is not a complete record). However, I can successfully harvest our thredds instance so maybe if I try yours we can at least get closer to troubleshooting whether it is GN 3.0 or Thredds misbehaving for you.
In case this helps, the settings I currently use are:
Catalog URL > http://xxx.xxx.xx/thredds/xxxx/xxx.xml
Use Account > not checked
X Create ISO19119 metadata for all services in the catalog
X Create metadata for Collection Datasets
X Ignore harvesting attribute
X Choose Extract DIF metadata elements and create ISO metadata
Select schema for output metadata records > iso19139
X Create metadata for Atomic Datasets
X Ignore harvesting attribute
X Choose Extract DIF metadata elements and create ISO metadata
Select schema for output metadata records > iso19139
NOTE: The .xml is important for the catalog URL and you do need to make sure the iso19139 stylesheet is loaded in your templates from the admin page (but it won't even show in the dropdown if it's not, so I'm guessing that is not your issue).
*More on just Thredds harvests :
I have been struggling with getting a good harvest of Thredds. It does weird things with the title and does not pull in all the metadata. This is a known issue and one of the developers on this list tells me it has to do with the fact that the current Geonetwork Thredds harvester was written before ncISO was integrated into Thredds. I've tried to play with rewriting the stylesheet but haven't had much success. HOWEVER, I can go to Thredds, save the ISO .xml page to my computer and do an 'insert metadata' using the gmiTogmd stylesheet. It's manual but it does pull in all the correct metadata and gives me hope that there might be a way to fix the automated part of it. You might want to try that manually and see if you can get the metadata into GN that way.
Kathy
-----Original Message-----
From: Alberto Brosich [mailto:abrosich@…1433…]
Sent: Wednesday, October 21, 2015 10:39 AM
To: geonetwork-users@lists.sourceforge.net
Subject: Re: [GeoNetwork-users] Thredds harvesting issue
Hi Filipe,
the catalog url is correct, I pasted in a browser to check it.
It points to a thredds server on localhost (http://localhost:8080/…).
I don't know if could be a problem.
I tried the same options as you (hence without "Create metadata for Atomic Datasets" option, but i will need it because my metadata are inside netcdf files) and the exception changed:
Thrown Exception java.lang.IllegalArgumentException: Schema not registered : during dataset processing
It seems like it doesn't use the schema I specified in the text box
("iso19139") and has a "null" schema.
Furthermore I cannot save the configuration of the harvester and I find the following exception in the log:
2015-10-21 16:00:48,810 ERROR [jeeves.service] - Exception when executing service
2015-10-21 16:00:48,810 ERROR [jeeves.service] - (C) Exc :
org.quartz.JobPersistenceException: The job
(HARVESTER_GROUP_NAME.1717897d-88ee-40b4-816c-867b35960d3b) referenced by the trigger does not exist.
2015-10-21 16:00:48,810 ERROR [jeeves] - Error occurred within a transaction
org.quartz.JobPersistenceException: The job
(HARVESTER_GROUP_NAME.1717897d-88ee-40b4-816c-867b35960d3b) referenced by the trigger does not exist.
But I suppose this is another problem therefore I will send another mail with an appropriate subject.
Regards
Alberto
<...snip....>
