Esri/geoportal-server-harvester

How to harvest a Entreprise Geodatabase

Closed this issue · 11 comments

Hi
In order to harvest an Entreprise Geodatabase, should I export all the metada into XML metadata files ?
Is It possible to create a connector ?

Best regards

In general, exporting XML or JSON from any source, storing it in a folder, then harvesting this folder is always a viable option. However, in your case it would be probably a very tedious job.

A better option would be to write a custom inbound adapter. In order to do that you would need to have an access to the ArcGIS Runtime SDK.

Hi
If you have access to an arcsde sdk (files : jpe_sdk.jar, jsde_sdk.jar) it should be possible to write an harvester. GeoNetwork has such an harvester.

https://geonetwork-opensource.org/manuals/2.10.4/eng/users/managing_metadata/harvesting/sde/index.html

that is correct. but similar to geonetwork, this would require additional installation, since we cannot ship those jar files with the harvester. similarly, the geonetwork license is not compatible with our geoportal license so it prevents us from combining their harvester into ours. open source/proprietary doesn't make a difference here. it's the license that dictates what you can do with some component.

In ArcMap we have a publishing client that pushes metadata to a geoportal (instead of a harvester pulling). Do you use ArcMap? if so, check out https://github.com/Esri/geoportal-server/tree/master/components/desktop/PublishClient

Is the 10.6 ArcCatalog Publishing Client compatible with Harvester version 2.5.2? I have a brand new GeoPortal Server/Harvester on the same server. I'm trying to populate it with XML's extracted from a geodatabase. When I follow the Publishing Client use instructions I get errors. I tried pointing it at both the /geoportal and the /harvester server URLs with no luck.

zguo commented

The ArcCatalog Publish Client is intended for publishing to geoportal 1.2.x, if your metadata from geodtabase is already extracted to a folder, you can use harvester to harvest the folder into a geoportal 2.x instance.

zguo commented

Implemented in the updated harvester, now you can harvest geodatabase metadata directly through JDBC database broker, please see https://github.com/Esri/geoportal-server-harvester/wiki/Brokers#jdbc-database.

I have followed the guidance in https://github.com/Esri/geoportal-server-harvester/wiki/Brokers#jdbc-database
Copied mssql-jdbc-10.2.0.jre11 into ..\Tomcat 9.0\lib
Added lines in POM.xml

com.microsoft.sqlserver
mssql-jdbc
10.2.0.jre11

Broker

Tried older version as well.
Unfortunately, I couldn’t make it work. The log file error is not clear to me.
Just wondering if there any other documentation (like step by step).

you should not have to rebuild the harvester for this. Just putting the JDBC driver in the Tomcat/lib folder, restarting Tomcat after this, and providing the correct info should be sufficient. Can you provide the error message?

in the fields for the select statement, primary key, title and description: does this match your table structure? is the database_name unique?

hmm... seems the log file did not get attached,

question:

  • can you sign in to the SQL Server (jdbc:sqlserver://desktop-ni994f7:1433;databaseName=EWR) from SQL Server Management Studio?
  • if so, since you sign in using account 'gisadmin' what is the default database for this user? do they have access to EWR?
  • if so, what is the structure of the table dbo.gdb_items?
  • is this an Esri Enterprise Geodatabase? if so:
    • use OBJECTID as primary key
    • use 'Name' as the title column
    • use 'Documentation' as the field with the metadata for feature classes