[ogsa-wg] OGSA-WG activities on the Express Profile

Hiro Kishimoto hiro.kishimoto at jp.fujitsu.com
Wed Oct 17 09:33:06 CDT 2007


Hi Duane,

Minor procedural comment.

> As for the status of the OGSA-SP within the OGF document process, the
> OGSA-SP is on schedule to enter public comment within the week.

I would say the OGSA-SP is on schedule to make WG final call in this
week and submit to the OGF editor for public comments in the following
week.

Thanks,
----
Hiro Kishimoto

-------- Original Message  --------
Subject: Re:OGSA-WG activities on the Express Profile
From: Duane Merrill III <dgm4d at virginia.edu>
To: Alan Sill <Alan.Sill at ttu.edu>, Andrew Grimshaw <grimshaw at virginia.edu>
Cc: "Alan Sill" <Alan.Sill at ttu.edu>, "Shannon Hastings"
<hastings at BMI.OSU.EDU>, "Hiro Kishimoto" <hiro.kishimoto at jp.fujitsu.com>
Date: 2007/10/17 6:39

> Shannon and Alan,
>  
> It sounds as if the caGrid /getServiceSecurityMetadata/() operation has 
> similar intent as the OGSA Security Profile (OGSA-SP, a.k.a., "Express 
> Profile") documents: to advertise and facilitate interoperable, secure 
> communication (or to cleanly determine if interoperability is not 
> possible). 
>  
> The mechanism for discovery is quite different, however.  In part, the 
> OGSA-SP documents are profiles on WS-Addressing: how to convey 
> secure-communication requirements within endpoint references (EPRs).  If 
> we were to apply the OGSA-SP  to your use-case below, the EPR used by a 
> client to interact with a resource would indicate that the endpoint 
> requires a TLS transport layer (possibly also indicating a requirement 
> for mutual-authentication at the TLS layer).
>  
> Why do this within the EPR?  The EPR datastructure plays a crucial role 
> for addressing resources in the OGSA architecture and is extensively 
> incorporated into many of its service interfaces (e.g., notification, 
> execution management, directory services, resource management, etc.).  
> The utility of the grid derives from the dynamic composition of 
> services, and EPRs are how services refer to and address each other.  
> They encapsulate the "invocation context" for a resource: everything one 
> needs to know to communicate with it.  Publishing security requirements 
> elsewhere (e.g., within WSDL, or exposed through reflective service 
> operations) incurs extra communication cost, availability and 
> interoperability risks, etc. 
>  
> How are communication requirements expressed?  The OGSA-SP documents use 
> the newly ratified OASIS WS-SecurityPolicy standard to convey 
> secure-communication requirements. Although resources are free to 
> express arbitrary security requirements, the OGSA-SP documents define 
> well-known, composable security policies for common security 
> mechanisms.  (And are perfectly extensible as new mechanisms are 
> introduced.)  For example, the OGSA-SP presently contains composable 
> policies for server-authenticated TLS, mutually-authenticated TLS, 
> UsernameToken authentication, mutual-X.509 authentication, digital 
> signature and encryption of document elements, etc., most of which defer 
> completely to sections of the WS-I Basic Security Profile.  In this 
> fashion, the OGSA-SP also serves as a location for minor mechanism 
> clarifications and refinements as Web services security mechanisms are 
> adapted to the Grid environment.
>  
> As for the status of the OGSA-SP within the OGF document process, the 
> OGSA-SP is on schedule to enter public comment within the week.  The 
> latest drafts of the OGSA-SP documents (and a potentially-languishing 
> use-case document) can be found at:
>  
> https://forge.gridforum.org/sf/docman/do/listDocuments/projects.ogsa-wg/docman.root.working_drafts.security_profiles_use_case
>  
> The OGSA-SP documents themselves are architected in a hierarchical, 
> extensible fashion.  The /OGSA-BSP Secure Addressing/ document defines 
> the general attachment mechanism for conveying security policy within 
> EPRs as well as how to make them tamper-resistant through digital 
> signature.  The /OGSA-SP Secure Transport/ document defines well-known 
> policies for de-facto transport-level secure communication 
> configurations.  The /OGSA-SP Secure SOAP Messaging/ document defines 
> analogous policies for de-facto message-level security mechanisms.
>  
> If you have someone at OGF and are interested in more detail, I will be 
> presenting the OGSA-SP documents at the
> /OGSA Express Authentication Security Profile/ session on Thursday 
> afternoon, 3:15 pm - 4:45 pm, and would enjoy discussing them further.
>  
> Cheers,
>  
> Duane
>  
>  
>  
>  
> ----- Original Message -----
> From: "Alan Sill" <Alan.Sill at ttu.edu <mailto:Alan.Sill at ttu.edu>>
> To: "Duane Merrill" <dgm4d at virginia.edu <mailto:dgm4d at virginia.edu>>; 
> "Andrew Grimshaw" <grimshaw at virginia.edu <mailto:grimshaw at virginia.edu>>
> Cc: "Alan Sill" <Alan.Sill at ttu.edu <mailto:Alan.Sill at ttu.edu>>; "Shannon 
> Hastings" <hastings at BMI.OSU.EDU <mailto:hastings at BMI.OSU.EDU>>; "Hiro 
> Kishimoto" <hiro.kishimoto at jp.fujitsu.com 
> <mailto:hiro.kishimoto at jp.fujitsu.com>>
> Sent: Tuesday, October 16, 2007 2:36 PM
> Subject: OGSA-WG activities on the Express Profile
> 
>  > Duane and Andrew,
>  >
>  > Can you catch us up, including Shannon, on the activities with 
>  > respect to the Express Profile and related documents, as they might 
>  > relate to the use case described below?  Shannon works with the 
>  > CaGrid group, and says that they have someone out at OGF who might be 
>  > able to interact with you about this if you are there.  Meanwhile, a 
>  > few web links and a status update as regards the use case documents 
>  > and draft profile might be useful.
>  >
>  > Thanks,
>  > Alan
>  >
>  > Begin forwarded message:
>  >
>  >> From: Shannon Hastings <hastings at BMI.OSU.EDU 
> <mailto:hastings at BMI.OSU.EDU>>
>  >> Date: October 16, 2007 12:43:22 PM CDT
>  >> To: CAGRID_USERS-L at LIST.NIH.GOV <mailto:CAGRID_USERS-L at LIST.NIH.GOV>
>  >> Subject: Re: operations added by Introduce
>  >> Reply-To: caGrid Users discussion Forum <CAGRID_USERS-L at LIST.NIH.GOV 
> <mailto:CAGRID_USERS-L at LIST.NIH.GOV>>
>  >>
>  >> Please do. We have recently become aware of another similar effort
>  >> within the grouper group.  Maybe this is the same effort.  Steve is 
>  >> out
>  >> at OGF this week so when he gets this email he can try to find those
>  >> folks as well.
>  >>
>  >> Thanks,
>  >> Shannon
>  >>
>  >>
>  >> Shannon Hastings
>  >> co-Director
>  >> Software Research Institute
>  >> Department of Biomedical Informatics
>  >> Ohio State University
>  >>
>  >> Office:  (614) 292-9461
>  >> Lab:  (614) 292-9845
>  >> hastings at bmi.osu.edu <mailto:hastings at bmi.osu.edu>
>  >>
>  >>
>  >> -----Original Message-----
>  >> From: caGrid Users discussion Forum [mailto:CAGRID_USERS-
>  >> L at LIST.NIH.GOV <mailto:L at LIST.NIH.GOV>]
>  >> On Behalf Of Alan Sill
>  >> Sent: Tuesday, October 16, 2007 12:27 PM
>  >> To: CAGRID_USERS-L at LIST.NIH.GOV <mailto:CAGRID_USERS-L at LIST.NIH.GOV>
>  >> Subject: Re: operations added by Introduce
>  >>
>  >> FYI, there is now an effort to define an architecture for
>  >> specification, advertising and discovery of security requirements of
>  >> exactly this nature in the context of the OGSA-WG "Express Profile"
>  >> as a profile under WS-Security.  If you are interested in this work,
>  >> I would be happy to put you in touch with members of that group so
>  >> you can read the profiles and use cases, and comment.
>  >>
>  >> Thanks,
>  >> Alan
>  >>
>  >> On Oct 16, 2007, at 10:48 AM, Shannon Hastings wrote:
>  >>
>  >>> To be even more specific just in case you are concerned about
>  >>> security.
>  >>>
>  >>> getServiceSecurityMetadata is an operation that is always public on a
>  >>> service.  What it does is tell any client what it required to talk to
>  >>> the service.  It contains no more information than that.  It 
>  >>> enables a
>  >>> client to know that he cannot talk to the service without a using TLS
>  >>> for example.
>  >>>
>  >>> The other three methods are part of the wsrf spec regarding resource
>  >>> property access.  These methods will use whatever security setting 
>  >>> you
>  >>> put on your service.  If you say that your service is secure and set
>  >>> authorization requirements at he service level in introduce, than
>  >>> these
>  >>> methods will also inherit this security.  So you have total 
>  >>> control of
>  >>> whether your resource properties are available.  Now keep in mind,
>  >>> there
>  >>> are resource properties of the service, not the service properties
>  >>> (which should be used for internal configuration and are not 
>  >>> available
>  >>> through any operations).
>  >>>
>  >>>
>  >>>
>  >>> Shannon Hastings
>  >>> co-Director
>  >>> Software Research Institute
>  >>> Department of Biomedical Informatics
>  >>> Ohio State University
>  >>>
>  >>> Office:  (614) 292-9461
>  >>> Lab:  (614) 292-9845
>  >>> hastings at bmi.osu.edu <mailto:hastings at bmi.osu.edu>
>  >>>
>  >>>
>  >>> -----Original Message-----
>  >>> From: caGrid Users discussion Forum [mailto:CAGRID_USERS-
>  >>> L at LIST.NIH.GOV <mailto:L at LIST.NIH.GOV>]
>  >>> On Behalf Of Shannon Hastings
>  >>> Sent: Tuesday, October 16, 2007 11:41 AM
>  >>> To: CAGRID_USERS-L at LIST.NIH.GOV <mailto:CAGRID_USERS-L at LIST.NIH.GOV>
>  >>> Subject: Re: operations added by Introduce
>  >>>
>  >>> These are required operations of any caGrid service.  They are 
>  >>> provide
>  >>> access to querying the resource properties of the main service as 
>  >>> well
>  >>> as the security configuration of the service.  This security
>  >>> information
>  >>> is just metadata about the requirements needed to connect to the
>  >>> service.  This enables a client to know that it will need a
>  >>> certificate,
>  >>> or will need to use SSL.  You should not disable these operations. 
>  >>> For
>  >>> more information on them.  The resource property ones are related
>  >>> to the
>  >>> WS specifications and the getSecurityMetadata is a caGrid only
>  >>> operation
>  >>> which you can read about at caGrid.org in the Introduce technical
>  >>> guide.
>  >>>
>  >>> Shannon
>  >>>
>  >>>
>  >>> Shannon Hastings
>  >>> co-Director
>  >>> Software Research Institute
>  >>> Department of Biomedical Informatics
>  >>> Ohio State University
>  >>>
>  >>> Office:  (614) 292-9461
>  >>> Lab:  (614) 292-9845
>  >>> hastings at bmi.osu.edu <mailto:hastings at bmi.osu.edu>
>  >>>
>  >>>
>  >>> -----Original Message-----
>  >>> From: caGrid Users discussion Forum [mailto:CAGRID_USERS-
>  >>> L at LIST.NIH.GOV <mailto:L at LIST.NIH.GOV>]
>  >>> On Behalf Of Joel Schneider
>  >>> Sent: Tuesday, October 16, 2007 11:27 AM
>  >>> To: CAGRID_USERS-L at LIST.NIH.GOV <mailto:CAGRID_USERS-L at LIST.NIH.GOV>
>  >>> Subject: operations added by Introduce
>  >>>
>  >>> I noticed that a service created using Introduce has the following
>  >>> four operations, in addition to the operations I defined by hand:
>  >>>
>  >>> getResourceProperty
>  >>> getMultipleResourceProperties
>  >>> queryResourceProperties
>  >>> getServiceSecurityMetadata
>  >>>
>  >>> This service in question was originally created using Introduce 
>  >>> 1.0 as
>  >>> an analytical service, and later migrated to Introduce 1.1.
>  >>>
>  >>> The presence of these additional operations raises a number of
>  >>> questions.
>  >>>
>  >>> 1) What is the purpose of the additional operations?
>  >>>
>  >>> 2) If these operations are not required by my service, is there an
>  >>> easy
>  >>>     or recommended way to remove or disable them?
>  >>>
>  >>> 3) From a security perspective, I'm concerned these operations might
>  >>>     expose the service's ServiceConfiguration to a remote user.  Is
>  >>> that
>  >>>     the case?  (Where is the code which implements these operations?)
>  >>>
>  >>> Help with these questions would be much appreciated.
>  >>>
>  >>> Best regards,
>  >>> Joel
>  >>>
>  >>> --
>  >>> Joel Schneider                     National Marrow Donor Program
>  >>> Software Developer (Contractor)    3001 Broadway Street NE
>  >>> phone: 612-617-8321                Minneapolis, MN 55413
>  >>> email: jschneid at nmdp.org <mailto:jschneid at nmdp.org>           
> http://www.marrow.org/
>  >
>  > Alan Sill, Ph.D
>  > TIGRE Senior Scientist, High Performance Computing Center
>  > Adjunct Professor of Physics
>  > TTU
>  >
>  > ====================================================================
>  > :  Alan Sill, Texas Tech University  Office: Admin 233, MS 4-1167  :
>  > :  e-mail: Alan.Sill at ttu.edu <mailto:Alan.Sill at ttu.edu>   ph. 
> 806-742-4350  fax 806-742-4358  :
>  > ====================================================================
>  >




More information about the ogsa-wg mailing list