Closed slead closed 8 years ago
You'll want to do this from the geoportal admin page. If you enabled the 'apply to all' option you would select the records based on the site they were harvested from. Then select the first page and select the 'apply to entire selection' option before selecting 'delete' from the drop down.
This apply to all option is set in gpt.xml.
After doing that, Pgadmin shows that those tables now contain 0 records. But searching within This Site still lists thousands of records:
Trying to view the Details of a record shows that it has been deleted:
So it appears that there's an index out of synch? Also C:\Lucene still contains ~50Mb of data, which could be the old records, perhaps?
I tried restarting Tomcat and restarting the whole PC but neither changes the fact that thousands of records are listed when running a search on This Site.
Any further clues? thanks
yes, the index will need to be updated. you can wait for the scheduled task to kick in. you should also be able to restart tomcat as that typically also starts the sync. obviously, you could remove the index files and then reapprove all remaining records. that will also typically result in rebuilding the index.
restarting Tomcat again didn't work, so I guess I'll wait for the scheduled task to kick in. Is there a way to determine when this will occur?
see the scheduler settings in gpt.xml. it allows you to set an interval for re-indexing, but you could also force it to happen at a certain time (say 5 minutes after now).
I accidentally imported thousands of records from http://gptogc.esri.com/geoportal while testing the CSW harvesting.
How can I quickly delete all records from the database so I can start afresh?
Will it cause any problems if I use PGADMIN and delete the records from _gptresource and _gpt_resourcedata? If that's a valid approach, do I need to create new indexes (etc) after doing this?
Thanks