I had a second so I finished this.  Trunk now has support for basic auth.
You enter the credentials on the server tab underneath the API
credentials.  Please give it a try and let me know if it works for you.

Karl



On Fri, May 24, 2013 at 11:28 AM, Karl Wright <[email protected]> wrote:

> CONNECTORS-692.  I will probably look at this over the weekend.
>
> Karl
>
>
> On Fri, May 24, 2013 at 11:26 AM, Karl Wright <[email protected]> wrote:
>
>> Hi TC,
>>
>> Unless I'm very much mistaken, there are no Apache kerberos session
>> cookies being used on your site, so it should be a straightforward matter
>> to include basic auth credentials to your Apache mod-auth-kerb module for
>> all pages during crawling.
>>
>> I'll create a ticket for this.
>>
>> Karl
>>
>>
>>
>> On Fri, May 24, 2013 at 11:14 AM, TC Tobin-Campbell <[email protected]> wrote:
>>
>>>  Hi Karl,****
>>>
>>> Here’s what I know so far.****
>>>
>>> ** **
>>>
>>> Our module is configured to use two auth methods: Negotiate and Basic.
>>> In most cases, we use Negotiate, but I’m guessing you’d prefer Basic.***
>>> *
>>>
>>> ** **
>>>
>>> Here’s an example header.****
>>>
>>> ** **
>>>
>>> GET / HTTP/1.1****
>>>
>>> Host: wiki.epic.com****
>>>
>>> User-Agent: Mozilla/5.0 (Windows NT 6.1; WOW64; rv:20.0) Gecko/20100101
>>> Firefox/20.0****
>>>
>>> Accept: text/html,application/xhtml+xml,application/xml;q=0.9,*/*;q=0.8*
>>> ***
>>>
>>> Accept-Language: en-US,en;q=0.5****
>>>
>>> Accept-Encoding: gzip, deflate****
>>>
>>> Cookie: wooTracker=QOMVLXDIC6OGOUXMGST1O54HYW573NNC;
>>> .EPICASPXAUTHQA=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;
>>> wiki_pensieve_session=j1pcf1746js1442m7p92hag9g1; wiki_pensieveUserID=5;
>>> wiki_pensieveUserName=Lziobro;
>>> wiki_pensieveToken=********************be3a3a990a8a****
>>>
>>> Connection: keep-alive****
>>>
>>> Authorization: Basic bHppb**************xMjM0   <-I've censored this
>>> line so you cannot get my password****
>>>
>>> ** **
>>>
>>> If I’m understanding you correctly, there’s no way to accomplish this
>>> currently? Or, is there some workaround we could implement? ****
>>>
>>> ** **
>>>
>>> *TC Tobin-Campbell *| Technical Services | Willow | *Epic*  | (608)
>>> 271-9000 ****
>>>
>>> ** **
>>>
>>> *From:* Karl Wright [mailto:[email protected]]
>>> *Sent:* Thursday, May 16, 2013 12:05 PM
>>> *To:* [email protected]
>>> *Subject:* Re: ManifoldCF and Kerberos/Basic Authentication****
>>>
>>> ** **
>>>
>>> Hi TC,
>>>
>>> Apparently mod-auth-kerb can be configured in a number of different
>>> ways.  But if yours will work with basic auth, we can just transmit the
>>> credentials each time.  It will be relatively slow because mod-auth-kerb
>>> will then need to talk to the kdc on each page fetch, but it should work.
>>> Better yet would be if Apache set a browser cookie containing your tickets,
>>> which it knew how to interpret if returned - but I don't see any Google
>>> evidence that mod-auth-kerb is capable of that.  But either of these two
>>> approaches we could readily implement.****
>>>
>>> FWIW, the standard way to work with kerberos is for you to actually have
>>> tickets already kinit'd and installed on your machine.  Your browser then
>>> picks up those tickets and transmits them to the Wiki server (I presume in
>>> a header that mod-auth-kerb knows about), and the kdc does not need to be
>>> involved.  But initializing that kind of ticket store, and managing the
>>> associated kinit requests when necessary, are beyond the scope of any
>>> connector we've so far done, so if we had to go that way, that would
>>> effectively make this proposal a Research Project.****
>>>
>>> What would be great to know in advance is how exactly your browser
>>> interacts with your Apache server.  Are you familiar with the process of
>>> getting a packet dump?  You'd use a tool like tcpdump (Unix) or wireshark
>>> (windows) in order to capture the packet traffic between a browser session
>>> and your Apache server, to see exactly what is happening.  Start by
>>> shutting down all your browser windows, so there is no in-memory state, and
>>> then start the capture and browse to a part of the wiki that is secured by
>>> mod-auth-kerb.  We'd want to see if cookies get set, or if any special
>>> headers get transmitted by your browser (other than the standard Basic Auth
>>> "Authentication" headers).  If the exchange is protected by SSL, then
>>> you'll have to use FireFox and use a plugin called LiveHeaders to see what
>>> is going on instead.****
>>>
>>> Please let me know what you find.****
>>>
>>> Karl****
>>>
>>> ** **
>>>
>>> ** **
>>>
>>> On Thu, May 16, 2013 at 12:37 PM, Karl Wright <[email protected]>
>>> wrote:****
>>>
>>> Hi TC,****
>>>
>>> Thanks, this is a big help in understanding your setup.****
>>>
>>> I don't know enough about exactly *how* mod-auth-kerb uses Basic Auth to
>>> communicate with the browser, and whether it expects the browser to cache
>>> the resulting tickets (in cookies?)  I will have to do some research and
>>> get back to you on that.****
>>>
>>> Basically, security for a Wiki is usually handled by the Wiki, but since
>>> you've put added auth in front of it by going through mod-auth-kerb, it's
>>> something that the Wiki connector would have to understand (and emulate
>>> your browser) in order to implement.  So it does not likely support this
>>> right now.  It may be relatively easy to do or it may be a challenge -
>>> we'll see.  I would also be somewhat concerned that it may not possible to
>>> actually reach the API urls through Apache; that would make everything moot
>>> if it were true.  Could you confirm that you can visit API urls through
>>> your Apache setup?****
>>>
>>> Karl****
>>>
>>> ** **
>>>
>>> On Thu, May 16, 2013 at 12:21 PM, TC Tobin-Campbell <[email protected]> wrote:
>>> ****
>>>
>>> Hi there,****
>>>
>>> I'm trying to connect ManifoldCF to an internal wiki at my company. The
>>> ManifoldCF wiki connector supplies a username and password field for the
>>> wiki api, however, at my company, a username and password is required to
>>> connect to the apache server running the wiki site, and after that
>>> authentication takes place, those credentials are passed on to the wiki api.
>>> ****
>>>
>>>  ****
>>>
>>> So, essentially, I need a way to have ManifoldCF pass my windows
>>> credentials on when trying to make its connection. Using the api login
>>> fields does not work.****
>>>
>>>  ****
>>>
>>> We use Kerberos the Kerberos Module for 
>>> Apache<http://modauthkerb.sourceforge.net/index.html>(AuthType Kerberos).  
>>> My understanding based on that linked documentation
>>> is that this module does use Basic Auth to communicate with the browser.
>>> ****
>>>
>>>  ****
>>>
>>> Is there anything we can to make ManifoldCF authenticate in this
>>> scenario? ****
>>>
>>>  ****
>>>
>>> Thanks,****
>>>
>>>  ****
>>>
>>>  ****
>>>
>>> *TC Tobin-Campbell *| Technical Services | Willow | *Epic*  | (608)
>>> 271-9000 ****
>>>
>>>  ****
>>>
>>> Sherlock <https://sherlock.epic.com/> (Issue tracking)****
>>>
>>> Analyst 
>>> Toolkits<https://sites.epic.com/epiclib/epicdoc/Pages/analyst/default.aspx>
>>> (Common setup and support tasks)****
>>>
>>> Report 
>>> Repository<https://documentation.epic.com/DataHandbook/Reports/ReportSearch>(Epic
>>>  reports documentation)
>>> ****
>>>
>>> Nova 
>>> <https://nova.epic.com/Login/GetOrg.aspx?returnUrl=%2fdefault.aspx>(Release 
>>> note management)
>>> ****
>>>
>>> Galaxy <https://documentation.epic.com/OnlineDoc/Documents.aspx> (Epic
>>> documentation)  ****
>>>
>>>  ****
>>>
>>> ** **
>>>
>>> ** **
>>>
>>
>>
>

Reply via email to