gklyne / admiral-jiscmrd

Automatically exported from code.google.com/p/admiral-jiscmrd
MIT License
0 stars 0 forks source link

Databank demonstration system creates manifest with wrong links #30

Closed GoogleCodeExporter closed 8 years ago

GoogleCodeExporter commented 8 years ago
What steps will reproduce the problem?
1. Demonstration system has two virtual machines and a browser running on a 
MacBook host.  Each machine has a NAT network interface for external access, 
and local-only network interface for a network that connects ADMIRAL, Databank 
and the desktop.
2. Dataset submission to databank mostly works (except see Issue 29)
3. But attempting to access link to resource in deposited datraset, no 
response, link has wrong IP address.  The address used is on the NATted 
network, but does not correspond to any actual machine present.

Is there a way to override Datrabank's choice of host in the link URIs?

Original issue reported on code.google.com by gk-goo...@ninebynine.org on 2 Mar 2011 at 5:52

GoogleCodeExporter commented 8 years ago

Original comment by gk-goo...@ninebynine.org on 2 Mar 2011 at 5:52

GoogleCodeExporter commented 8 years ago
I tried adding a ServerName directive to the databank-wsgi apache 
configuration, but no joy.

Original comment by gk-goo...@ninebynine.org on 2 Mar 2011 at 6:05

GoogleCodeExporter commented 8 years ago
The uri root for the manifest is set in production.ini. Change the value in 
granary.uri_root to a uri of your choice. 

The best thing to do would be to:
1. Delete existing silos
2. Stop Apache
3. Edit production.ini
4. Start Apache

Now create new silos. The new datasets would be created with the new uri

If you cannot delete any of the existing silos:
1. Stop apache
2. Eit production.ini with the new uri
3. Edit the uri in the file persisted_state.json located in the silo's 
directory. Do this for all of the silos.
4. Edit manifest.rdf for each version of the dataset and change the uri to the 
new one. Do this for all datasets in all the silos. Edit the files directly on 
the hard disk where the silo is stored (do not use the web interface to do 
this). 
5. Start apache

Regards,
Anusha

Original comment by anus...@gmail.com on 2 Mar 2011 at 6:26

GoogleCodeExporter commented 8 years ago
Anusha - thanks.  I'll try this and close the issue if it works.

Original comment by bhavana....@gmail.com on 3 Mar 2011 at 4:34

GoogleCodeExporter commented 8 years ago
To fix this problem, I
1. deleted all datasets
2. stop apache
3. edit production.ini
4. edited silos/admiral-demo/persisted_state.json
5. edited silos/admiral_demo/pairtree-prefix.??
6. restarted Apache

Newly created datasets have correct values in manifest.  I haven't tested newly 
created silos, but for now I'll assume that works.

Original comment by gk-goo...@ninebynine.org on 4 Mar 2011 at 9:59