- ogsa-hpcp-wg - lists.ogf.org

WSDL and schema files
by Christopher Smith 26 Sep '06

26 Sep '06

Hey there, Does anybody have the WSDL and schema files corresponding to the latest BES spec? I've attached the HPCProfileApplication schema, so you don't have to extract it from the appendix in the specification. -- Chris

1 0

New version of HPC Profile Application extension
by Christopher Smith 26 Sep '06

26 Sep '06

Hi all, I've uploaded a new draft of the HPC Profile Application extension ... this time to the JSDL-WG gridforge site. I've made changes to the document according to resolved issues, and then added some text to address the path name syntax issues. Those I have left as version tracked changes in the document so that people can easily see the additions and review them. Todo: - add normative schema in appendix - any format beautification to reflect OGF doc formats -- Chris

1 1

Re: [ogsa-hpcp-wg] [Fwd: SC06 Demo Kickoff]
by Dave Berry 25 Sep '06

25 Sep '06

Would someone care to fill me in on these plans for an SC06 demo? Dave. -----Original Message----- From: ogsa-hpcp-wg-bounces(a)ogf.org [mailto:ogsa-hpcp-wg-bounces@ogf.org] On Behalf Of Peter G. Lane Sent: 22 September 2006 18:50 To: 'Mailing List for OGSA-HPCP-WG' Subject: [ogsa-hpcp-wg] [Fwd: SC06 Demo Kickoff]

3 2

[SC2006 Demo] informal minutes from friday's OGSA HPC Profile SC 2006 interop minutes
by Marty Humphrey 25 Sep '06

25 Sep '06

Folks, These are off the top of my head, so apologies if I miss anything/anyone. Attendees: Marty Humphrey, UVA; Glenn Wasson, UVA; Duane Merrill, UVA; Mark Morgan, UVA; Dave Snelling, Fujitsu; Chris Smith, Platform; Rich Ciapala, Microsoft; Peter Lane, Globus We first agreed that this mailing list (OGSA HPC Profile) would be used for communications regarding the demo and/or interop in general. No separate mailing list is warranted. We then agreed that we would move this call FROM Fri 2pm EDT to Fri 11am EDT from this point forward, to better support our colleagues from Europe (note: as of now, we don't have any Asian-Pacific participants, who might feel a negative impact from moving this call). Topic 1: Security. After a good discussion, we decided that everyone would implement "WS-Security Username Profile (plaintext)", server-only-authentication HTTPS, and maybe client-side certificates if time permits. Consensus was that this is an excellent and pragmatic choice. We did not directly discuss authorization because of lack of time on this call. Topic 2: WS-Addressing. We agreed to use the W3C recommendation: 2005-08 namespace. This is a good forward-looking choice, and spreads the pain around fairly evenly. Topic 3: Deliverables and testing. We agreed that we would keep a web page to keep track of progress. The URL is: http://www.cs.virginia.edu/~humphrey/SC2006.htm This page will ULTIMATELY (soon?) become password-protected. The current content just captures the general flow of the process and information for the SC 2006 demo. This page content and style will be refined in the coming weeks. This version right now reflects the nuts-and-bolts" nature of the immediate job before us; the final version of the demo is the subject of future calls. Everyone agreed that this was an excellent and very productive kick-off meeting! Additional participants are greatly encouraged to participate. Please contact me directly or submit email to the mailing list. -- Marty Marty Humphrey Assistant Professor Department of Computer Science University of Virginia

3 3

[Fwd: SC06 Demo Kickoff]
by Peter G. Lane 22 Sep '06

22 Sep '06

1 0

SC06 Demo Kickoff
by Peter G. Lane 22 Sep '06

22 Sep '06

Forwarded from private emails... [From: Marty Humphrey <humphrey(a)cs.virginia.edu>] > Toll Free Number: 866-500-6738 > International: 1-203-480-8000 > Passcode: 231867 > > 2pm EDT today > [From: David Snelling <David.Snelling(a)UI.Fujitsu.com>] > Now to the point: There has been some chat about security among some of us. I will try to summarize what the options are and we can discuss them on a call. > > 1) No security: We use basics http without any protection. This would work fin in a closed setting, but interops work much better when we can test them remotely before hand. Not all of us will be allowed to publish unprotected endpoints. > > FLE View: We can do this, but only with the client also on the phone. Not a preferred option. > > 2) Use Mutually Authenticated HTTPS: This is effectively the "OGSA Basic Security Profile 1.0 - Secure Channel" specification. It assumes that both client and service have the root CA cert of the other's certificate. This is used to set up the SSL connection and the identity extracted from connection and used for authorization and account mapping. I believe this can be made to work with primary certs or with proxy certs. > > FLE View: This is our preferred option (with primary certs, but I think we can make proxies work with some effort). > > 3) WS-Security User Name Profile: As I understand it this encodes a user name (and possibly password) in a SOAP header element. The underlying transport is assumed to be HTTPS, because I don't believe that the message itself is encrypted. > > FLE View: We don't implement this at present and have no plans to do so. User names are not very Grid-like and so not on our roadmap. > > 4) WS-Security X.509 Profile: As I understand it this encodes the user's X.509 certificate in a SOAP header element. As above, the underlying transport is assumed to be HTTPS, and the message is not encrypted, although this is possible within this profile. This actual profile is very complicated and we would need to define very carefully which parts of it's many optional bits we use. > > FLE View: We don't implement this at present, but would be willing to do so. However, this would put our participation at risk depending on how complicated a profile the group chooses and how soon we decide. > > I look forward to a call. > -- > Take care: > > Dr. David Snelling < David . Snelling . UK . Fujitsu . com > > Fujitsu Laboratories of Europe > Hayes Park Central > Hayes End Road > Hayes, Middlesex UB4 8FE > > +44-208-606-4649 (Office) > +44-208-606-4539 (Fax) > +44-7768-807526 (Mobile)

1 0

Proposed rephrasing of the restriction on BES vector operations in the HPC Profile
by Marvin Theimer 18 Sep '06

18 Sep '06

Hi; As you may recall from the discussions at the Tokyo GGF conference, it was agreed that the HPC Profile base case would be kept as simple as possible in order to make it as easy as possible to create a compliant BES service (and to forestall feature creep). Among other things, that meant that we agreed upon requiring the HPC Profile-compliant BES services would only be required to support BES vector operations of length 1. But of course we'd like to allow compliant BES services to support vectors of length greater than 1 if they so desire (which we expect to be the common case). Enclosed is a rewording of the paragraph in section 4 of the HPC Profile specification draft that restricts vector lengths of BES operations. I've rewritten it to clarify that vector lengths greater than 1 are allowed but may result in a fault. Please let me know if you would like additional rewording to occur. If I don't hear any rewording suggestions in the next few days then I'll go ahead and insert the rewording shown into the HPC Profile document. Reworded paragraph: The BES GetActivitiesStatus, TerminateActivities, and GetJSDLDocuments operations include a vector input parameter that specifies the set of activities that the operation should be applied to. A Compliant system MUST be prepared to support a vector length of 1. A Compliant system SHOULD support input vector lengths greater than 1 but MAY return a Not_Supported fault in response to input vector lengths greater than 1. Thanks, Marvin.

5 5

Brief HPC Profile teleconf. call on Friday at the usual time
by Marvin Theimer 15 Sep '06

15 Sep '06

Hi; I think this may be a short conference call. The main agenda item I propose we discuss is next-steps for interop plans among those parties that are interested. Date: Friday, 9/8/06 Time: 10AM EDT (7AM PDT) Connect information: 1-866-500-6738 (from US & Canada) or 1-203-480-8000 Passcode: 414088 Marvin.

3 2

Type for file elements in HPC Profile Application extension
by Christopher Smith 14 Sep '06

14 Sep '06

During the HPC Profile working group session yesterday (after JSDL session 1), Mark Morgan expressed an opinion that the file elements for input/output/error should be left as strings rather than changed to anyURI. The idea is that either the back end system accepts the string, or doesn't, and we don't put any constraints on the format. The description of those elements would then contain some statement that the implementation should return a particular kind of fault if it doesn't accept the provided string. I tend to like this approach purely from the laziness position that I don't have to craft some statements about exactly which formats are supported, blah, blah, blah. Mark also mentioned that it greatly eases implementation. There are some downsides from the interop point of view, but I think those exist anyway, as some systems might support some form of URIs, whereas others would not, etc, etc. So ... I would like to propose to keep the type of these elements as a string. Comments? Mark ... please chime in if I've misrepresented you in any way. -- Chris

1 0

Nominations sought for co-chair of HPC Profile WG
by Marty Humphrey 14 Sep '06

14 Sep '06

Folks, As we discussed in the HPC Profile session yesterday, Marvin Theimer must step down as the co-chair of the HPC Profile WG because of his new job responsibilities. (Note that I have been assured that this by no means implies that Microsoft will reduce or eliminate their effort in our WG). Marvin is a tremendous guy and has done a FANTASTIC job as the co-chair and we will miss his energy, experience, and vision. As a group, we have a number of things that we still need to push through, namely the finalization of the three documents of the group as well as our emerging interop/compliance suite efforts. Please send me email if you would like to nominate yourself or someone else as the co-chair of the HPC Profile group. The deadline for submissions is Friday Sept 22 (1 week from today). Thanks, Marty Marty Humphrey Assistant Professor Department of Computer Science University of Virginia

1 0