Hi Peter,
The LCF solr output connector uses the Solr http API to post in documents and metadata into the Solr pipeline and index.  What I meant by that was that I will modify the Solr output connector to also transmit the access tokens to the Solr http API as additional document metadata.

From: ext Peter Sturge [mailto:peter.sturge@googlemail.com]
Sent: Tuesday, April 20, 2010 6:51 PM
To: connectors-user@incubator.apache.org
Subject: Re: FW: Solr and LCF security at query time

Hi Karl,

Thanks for the info. I'll have a look at the link and try to take in as much sugar as my insulin levels will handle...
It sounds like the necessary interface(s) are already in LCF - just a matter of implementing them in the Solr 1872 plugin.
I'll need to digest the LCF stuff to get to grips with it..please bear with me while I do that...

When you say:
   The LCF solr output connection doesn't yet do this, but it is trivial for me to make that happen.
Do you mean a mechanism by which solr.war can get url et al info from its parent container (Tomcat, Jetty etc.), or have I misinterpreted this?


On Tue, Apr 20, 2010 at 11:05 PM, <karl.wright@nokia.com> wrote:
Hi Peter,
I'm the principal committer for LCF, but I don't know as much about Solr as I ought to, so it sounds like a potentially productive collaboration.
LCF does exactly what you are looking for - the only issue at all is that you need to fetch a URL from a webapp to get what you are looking for.  The "plugs" are all inside LCF for different kinds of repositories.  Here's a link that might help with drinking the LCF "koolaid", as it were: https://cwiki.apache.org/confluence/display/CONNECTORS/Lucene+Connectors+Framework+concepts
The url would be something like this (on a locally installed tomcat-based LCF instance):
... and this fetch returns something like:
... which represent the amalgamated tokens for all of the defined authorities, and by some strange coincidence ( ;-) ) are compatible with certain pieces of metadata that have been passed into Solr with each document - one set of Allow tokens, and a second set of Deny tokens.  The LCF solr output connection doesn't yet do this, but it is trivial for me to make that happen.
Does this sound plausible to you?


From: ext Peter Sturge [mailto:peter.sturge@googlemail.com]
Sent: Tuesday, April 20, 2010 5:41 PM
To: connectors-user@incubator.apache.org; dev@lucene.apache.org

Subject: Re: FW: Solr and LCF security at query time

Hi Karl,

Integrating LCF to get external token support for SOLR-1872 sounds very interesting indeed. I don't know anything about LCF, but one of the things I was planning for SOLR-1872 is to make acl.xml (or rather its behaviour) 'pluggable' - i.e. it would just be one of a series of plugins that could be used for obtaining back-end authentication information.

If you're good with LCF, perhaps we could work together to build this in. One of the first things would be defining an interface that would be as easy as possible to plug LCF into. Have you any suggestions/insight on this front?

Many thanks,

On Tue, Apr 20, 2010 at 4:08 PM, <karl.wright@nokia.com> wrote:
SOLR-1872 looks exactly like what I was envisioning, from the search query perspective, although instead of the acl xml file you specify LCF stipulates you would dynamically query the lcf-authority-service servlet for the access tokens themselves.  That would get you support for AD, Documentum, LiveLink, Meridio, and Memex for free. It seems likely that this component could be modified to work with LCF with minor effort.
The missing component still seems to be AD authentication, which needs a solution.

From: ext Peter Sturge [mailto:peter.sturge@googlemail.com]
Sent: Tuesday, April 20, 2010 10:44 AM
To: dev@lucene.apache.org
Subject: Re: FW: Solr and LCF security at query time

If you want to do this completely within Solr, have a look at:
SOLR-1834 and SOLR-1872. These use a SearchComponent plugin for Solr.


On Tue, Apr 20, 2010 at 1:25 PM, <karl.wright@nokia.com> wrote:

From: Wright Karl (Nokia-S/Cambridge)
Sent: Tuesday, April 20, 2010 8:16 AM
To: 'dominique.bejean@eolya.fr'
Cc: 'solr-dev@apache.org'; 'connectors-dev@incubator.apache.org'; 'connectors-user@incubator.apache.org'
Subject: RE: Solr and LCF security at query time

Yes, I am aware of this ticket and contribution.  Luckily LCF establishes a powerful multi-repository security model, even though it doesn't yet do the final step of enforcing that model at the search end.  LCF allows you to define multiple authorities to operate against disparate repositories, and use the appropriate authority to secure any given document.  The solr people are aware of this design, which addresses the issues raised by SOLR-1834 very nicely.  However, as I said before, time is a problem, and the work still needs to be done.
I suggest you read up on the actual security model of LCF, and perhaps experiment with that and the SOLR-1834 contribution, to see if there is common ground.  One thing we've learned at MetaCarta is that post-filtering for security purposes is expensive, and it is better to modify the queries themselves to restrict the results, if possible.  I'm not sure which approach SOLR-1834 takes, although it sounds like it might be the filtering approach.  Still, it would be better than nothing.
Please let me know what you find out.

From: ext Dominique Bejean [mailto:dominique.bejean@eolya.fr]
Sent: Tuesday, April 20, 2010 8:03 AM
To: Wright Karl (Nokia-S/Cambridge)
Cc: connectors-user@incubator.apache.org; connectors-dev@incubator.apache.org
Subject: Re: Solr and LCF security at query time


Thank you for your reply.

I made some research today and I found this :

Sorl security model have to be able to filter result list with items coming from various sources at the same time (livelink, documentum, file system, ...). Big subject :)


Le 20/04/10 13:34, karl.wright@nokia.com a écrit :
Hi Dominique,
At the moment, in order to enforce the LCF security model within Lucene/Solr, you will need to build this functionality into whatever client you are using to display the Lucene search results.  Specifically, you would need to take the following steps:
(1) Have your users access your search client through Apache.
(2) Use the Apache module mod_auth_kerb, combined with LCF's mod_authz_annotate, to cause authorization HTTP headers to be transmitted to the client webapp.
(3) Have your client webapp alter whatever queries it is doing, to add an appropriate query clause for each of the access tokens transmitted in the headers.
(This is how it is done at MetaCarta.)
Alternatively, you may find a way to do this completely with a web application under a Java app server such as Tomcat.  I have not yet done the research to find out whether this is a feasible alternative.  Effectively, what you need something like mod_auth_kerb to do is to authenticate your user against Active Directory, or whomever the authenticator ought to be.  JAAS may be helpful here.
There are, of course, intentions to fill out the missing pieces more completely and transparently via a Solr search plugin and/or filter.  What has been lacking is time.  If you are in a position to do development in this area, we're happy to have any assistance you might provide. 

From: ext Dominique Bejean [mailto:dominique.bejean@eolya.fr]
Sent: Tuesday, April 20, 2010 5:06 AM
To: connectors-user@incubator.apache.org
Subject: Solr and LCF security at query time


I don't see in LCF wiki how Solr and LCF works together at query time in order to remove from the result list the items the user is not allowed to access.

In http://cwiki.apache.org/CONNECTORS/lucene-connectors-framework-concepts.html, I just see these sentences :

" Once all these documents and their access tokens are handed to the search engine, it is the search engine's job to enforce security by excluding inappropriate documents from the search results. For Lucene, this infrastructure is expected to be built on top of Lucene's generic metadata abilities, but has not been implemented at this time."

I am not sure to understand. Does this mean that for the moment, it is not possible for Solr to apply security by using an Authority Connector ?