ESAVO Full Harvestable Registry

[KevinBenson]

Okay yes the earliestDatestamp is just a small config change, I will do that
in the morning.  Let me know what you saw with a 0.9 namespace they should
all be 0.10.

As far as harvesting Esa well it might have to wait till the next astrogrid
release because I don't put a set=ivo_managed on the url.  I attempted to do
a harvest, but it failed with a 502 status code from the esa site (this is
using the automatic harvester).  I suspect it is trying to obtain everything
from your registry which causes it to time out.  So will need to wait till I
put the set piece on the automatic harvester.  But I have a manual way of
placing them in so I will grab your ivo_managed set and put them on the main
Leicester registry of galahad tomorrow.

cheers,
Kevin



-----Original Message-----
From: owner-registry at eso.org [mailto:owner-registry at eso.org]On Behalf Of
Aurelien Stebe
Sent: 25 July 2005 14:48
To: registry at ivoa.net
Cc: Pedro Osuna
Subject: Re: ESAVO Full Harvestable Registry


Thank you, Kevin.

I actually implemented the OAI interface returning resources with the
Registry Interface namespace
before realising that nearly all the registries were using the
VOResource v0.10 namespace.
I think it should be the RI one just as for the search interface, but
let's see what everybody thinks of this.

I'm having just two small problems with the Astrogrid registries for
harvest.
The missing sets is not a problem actually. The system looks for the
"ivo_managed"
set and if it can't find it, just assumes that the results returned are
the managed ones.
It is just that before harvesting, it checks for a few values, and the
"ivo_vor" metadataPrefix
namespace returned by Astrogrid registries is v0.9 although the
resources returned seem to be of v0.10.
Also, the earliestDatestamp is "undefined". I was hopping this would be
small config adjustements on your part.

Tell me if you have any problem harvesting us.

Cheers,
Aurelien


KevinBenson wrote:

>Very nice Aurelien, looks good.
>
>I will put it as one of the registries to be harvested at Astrogrid on
>Monday.  Sorry for being late on the next Interface Spec document, I should
>have it ready by Monday afternoon for review.
>
>Yes the Resource namespace after "metadata"  being the Registry Interface
is
>fine for Astrogrid registries, we tend to xsl what is below the "metadata"
>element to what is needed by our registries.  We might want to put that
>issue back on the list of discussions though.  Originally we had it as
>"Registry Interface" namespace as well, but early on registries kept having
>trouble harvesting us, so eventually decided to make things easier and xsl
>to where it is now vr:Resource.
>
>Let me know what problems your having with any Astrogrid Registries, I
>suspect it might be because we currently don't support sets.  So a
>ListRecords call is by default ivo_managed and does not support the "set"
>parameter to it.  I hope to support sets soon, but suspect another good 2
>weeks till it is on the production servers.
>
>Here are the known OAI urls to Astrogrid registries:
>http://galahad.star.le.ac.uk:8080/galahad-registry/OAIHandlerv0_10   //our
>main full+publishing leicester registry
>http://msslxv.mssl.ucl.ac.uk:8080/astrogrid-registry/OAIHandlerv0_10
>//publishing registry
>http://astrogrid.ast.cam.ac.uk/CASU-registry/OAIHandlerv0_10 //publishing
>registry --currently down at the moment
>http://hydra.star.le.ac.uk:8080/astrogrid-registry/OAIHandlerv0_10 //a
>full+publishign leicester registry that will disappear in the coming
months.
>//finally there is another full registry at Cambridge but it does not
really
>manage anything so nothing to harvest from that one.
>
>
>cheers,
>Kevin
>
>
>
>