August 2005 - ogsa-wg - lists.ogf.org

RE: [ogsa-wg] Document on resource models
by Subramaniam, Ravi 08 Aug '05

08 Aug '05

Hi, A thought crossed my mind after we dropped off from the call. With reference to CIM and Grid, what is the status of the Grid Schema WG? As I understood it (I think from Carl K) they were to apply CIM to execution and monitoring of Grid jobs. Maybe they have some useful stuff that we can "proudly plagiarize" for the F2F discussion. Thanks! Ravi -----Original Message----- From: owner-ogsa-wg(a)ggf.org [mailto:owner-ogsa-wg@ggf.org] On Behalf Of Hiro Kishimoto Sent: Sunday, August 07, 2005 7:14 AM To: Fred Maciel Cc: ogsa-wg(a)ggf.org Subject: Re: [ogsa-wg] Document on resource models Hi Fred, Andreas and I review your document and we agree with overall conclusion and recommendations. But we think that the output should be more concrete on how OGSA can use CIM. Please check out the attached slides and let discuss it on upcoming call. Thanks, ---- Hiro Kishimoto Fred Maciel wrote: > Hi, > > Here is the current version of the document I'm writing on resource > models. > I was assuming that the contents of this document would be the basis > for the discussions in the face-to-face, but I have been getting mixed > signals. In the teleconference I'd like to discuss two things: > > - What are we discussing in the one day that we allocated for resource > models in the face-to-face? Are we discussing resource models on execution > management, are we going to make a decision on which resource model we will > use, both, or something else? (I thought that we were going to make a > decision on what we will use). > > - If we are going to discuss what we will use, is there anything missing in > the document that you would like to see written? > > Regards, > > Fred Maciel > Hitachi America R&D >

1 0

BES Agenda at OGSA F2F
by Steven Newhouse 08 Aug '05

08 Aug '05

As discussed at the last call the provisional agenda for the BES part of the F2F (Monday 10.30-15.00) 0. Agenda bashing [10 mins] 1. Review of current draft & discussion [50 mins] 2. Discussion about renderings [60 + 60 mins] 3. Discussion about implementations [30 mins] Consider this a stake in the ground! Steven -- ---------------------------------------------------------------- Dr Steven Newhouse Tel:+44 (0)2380 598789 Deputy Director, Open Middleware Infrastructure Institute (OMII) Suite 6005, Faraday Building (B21), Highfield Campus, Southampton University, Highfield, Southampton, SO17 1BJ, UK

1 0

roadmap document status
by Hiro Kishimoto 08 Aug '05

08 Aug '05

Hi all, Greg Newby, GGF Editor, returns the roadmap document to OGSA-WG for final edit. > Status has changed: - New: 'Author Reviewing Edits' Old: 'Public Comment Period' > Resolution has changed: - New: 'Returned to Authors/Group' Old: 'Accepted' > Comment By: Greg Newby (2005-08-07 17:45:36) > Public comment for this document is completed. > Authors, please see the comments here: > https://forge.gridforum.org/forum/forum.php?forum_id=540 > > Provide any responses, and if desired upload a final > document for review. Please indicate your readiness > for final editor review in this tracker. -- Hiro Kishimoto

1 0

Proposed agenda for Aug. 8th call
by Hiro Kishimoto 08 Aug '05

08 Aug '05

Hi all, The following is a proposed agenda for OGSA-WG telecon on Aug. 8th Monday from 5pm to 7pm (CDT). The dial-in number for Monday; Free: +1-866-639-4741 Toll: +1-574-935-6703 PIN: 8980700 Screen share service will be provided. URL: http://ogsa.glance.net Session key: 0808 See more explanation: http://www-unix.gridforum.org/mail_archive/ogsa-wg/2004/06/msg00077.html 1) Early discussion (20 min) Note taker assignment Roll call Telecon minutes approval https://forge.gridforum.org/projects/ogsa-wg/document/minutes-20050803/en/1 Agenda bashing 2) August F2F meeting update (20 min) Fellow WG co-chairs: Please provide detailed agenda for your session. The latest meeting agenda. https://forge.gridforum.org/projects/ogsa-wg/document/OGSA-2005-Aug-F2F-age… 3) Provisioning discussion (20 min) CDDLM-WG and sub group of EMS team had joint call on Aug. 3. They wrote up the following agenda for F2F > A) (1:30-1:45) Introduction (learn more about EMS broader picture, in > the past wee overviewed CDDLM) > B) (1:45-3:15) Defining architectural relationship (This is really > related to the pictures we discussed above, Fig 5 in the OGSA arch & Fig > 7 in CDDLM Foundation) > C) (3:15-3:45) Indebt overview of the CDDLM interfaces (if needed) > D) (3:45-4:15) Description and integration of ACS & CDDLM (Ravi: added > BES as well). > E) (4:15-4:45) Integration roadmap > F) (4:45-5:00) Next steps 4) Resource Modeling discussion (60 min) > - Fred has started circulating a first draft to a small group of > people. It will be reviewed Monday to check that it addresses > the right area but detailed discussion on the content will be > left for the F2F. http://www-unix.gridforum.org/mail_archive/ogsa-wg/2005/08/msg00030.html 5) Wrap up AOB Aug. 10 (Wed) OGSA-AuthZ-WG, roadmap, Architecture and Glossary 1.5 Aug. 15-19 F2F meeting in SF bay area https://forge.gridforum.org/projects/ogsa-wg/document/telecon_schedule/ ---- Hiro Kishimoto

2 1

Re: [gfs-wg] Minutes of 8/5,6 GFS-WG Conference call
by Manuel Pereira 08 Aug '05

08 Aug '05

Hello all, Thank you Osamu for posting the minutes of the last meeting. I am sending this note to each of the associated mailing lists to notify all interested parties that the most recent RNS specification draft has been posted to GridForge and is available for review.[1] I would also like to inform everyone interested that we are planning to submit this draft to the GFD Editor this week. [1] https://forge.gridforum.org/projects/gfs-wg/document/RNS-Proposed_Final_Dra… Best regards, Manuel Pereira ------------------------------------------------------ IBM Almaden Research Center 1-408-927-1935 [Tieline: 457] mpereira(a)us.ibm.com ------------------------------------------------------ owner-gfs-wg(a)ggf.org wrote on 08/05/2005 06:50:23 PM: > Hi all, > > Here is minutes of the last conference call. > > Thanks, > Osamu > > --- > GFS-WG conference call > Friday, Aug 5, 2005 > 4:30pm to 5:30pm (PDT) > > Note taker: Osamu Tatebe (AIST) > > Participants > ------------ > Manuel Pereira > Ted Anderson > Osamu Tatebe > > Agenda > ------ > Introduction & agenda bushing ( 5min) > Change from 8th to 9th revision ( 5min) > Full path vs absolute path (20min) > Other issues (10min) > > Meeting > ------- > * Change from 8th to 9th revision > https://forge.gridforum.org/projects/gfs-wg/document/RNS- > Proposed_Final_Draft-v1.9/en/9 > > - make XML namespace in WSDL consistent with the specification > (reflecting Takuya Ishibashi's comment) > > - modify the last part of the introduction section caused by removing > the appendix of Grid File System Profile. > > - 'absolute path' in the second table in Section 1.3.3.1 "move", is > changed to 'full path'. > > * Full path vs absolute path > > - In Section 2.2.1.1 "Referral Messages", the first sentence states > that "all service operations [MUST] accommodate the use of absolute > pathnames," although all input to the service interface is now a > full path. > > - need to accept both a full path and an absolute path for all input > to the service. How to differenciate these two? > > - An absolute path starts with '/', and a full path does not start > with '/'. > > * Other issues > > - Can '.' and '..' be specified as a 'Name'? > > -> Yes. At least, RNS does not need such a restriction. > > FYI: NFS version 4 assigns no special semantics to the names "." and > "..". The LOOKUPP operator must be used to lookup a parent > directory. > > --- > Osamu Tatebe, Ph.D. Tel: +81-29-861-5844 FAX: +81-29-862-6601 > Grid Technology Research Center, > National Institute of Advanced Industrial Science and Technology (AIST) > AIST Tsukuba Central 2, 1-1-1 Umezono, Tsukuba, Ibaraki 3058568 JAPAN > E-mail: o.tatebe(a)aist.go.jp http://phase.hpcc.jp/people/tatebe/ >

1 0

Document on resource models
by Fred Maciel 07 Aug '05

07 Aug '05

Hi, Here is the current version of the document I'm writing on resource models. I was assuming that the contents of this document would be the basis for the discussions in the face-to-face, but I have been getting mixed signals. In the teleconference I'd like to discuss two things: - What are we discussing in the one day that we allocated for resource models in the face-to-face? Are we discussing resource models on execution management, are we going to make a decision on which resource model we will use, both, or something else? (I thought that we were going to make a decision on what we will use). - If we are going to discuss what we will use, is there anything missing in the document that you would like to see written? Regards, Fred Maciel Hitachi America R&D

2 1

RE: [ogsa-wg] Re: [cddlm] Notes from the OGSA EMS & CDDLM meeting
by Ziu, Peter 07 Aug '05

07 Aug '05

Hiro, I thought it might be useful to push up some concept visuals for autonomic/automated provisioning, that may assist in generating some thought and discussion at the face2face. Pete. -----Original Message----- From: owner-ogsa-wg(a)ggf.org [mailto:owner-ogsa-wg@ggf.org] On Behalf Of Hiro Kishimoto Sent: Wednesday, August 03, 2005 7:35 PM To: Milojicic, Dejan S Cc: cddlm-wg(a)ggf.org; ogsa-wg(a)ggf.org Subject: [ogsa-wg] Re: [cddlm] Notes from the OGSA EMS & CDDLM meeting Thanks Dejan for the meeting notes, I've put proposed joint session agenda into F2F schedule and update attendees list accordingly. https://forge.gridforum.org/projects/ogsa-wg/document/OGSA-2005-Aug-F2F- agenda If you have any farther questions/comments, please let me know. ---- Hiro Kishimoto Milojicic, Dejan S wrote: > By phone, 8/3, 2pm-3pm PST. > > Attendees: Mark Morgan, Hiro Kishimoto, Chris Smith, Stuart Schaefer, > Jun Tatemura, Ravi Subramanian, Mike Behrens, Steve Loughran, and > Dejan Milojicic. > > Notes taker: Dejan Milojicic. > > Agenda: > > 1. Go over the two documents (CDDLM Foundation document for EMS people > an the OGSA architecture document for the CDDLM people. > > Dejan suggested that CDDLM is part of the EMS, but there are other > pieces which are out of scope. > > Chris asked about activity v. solution. The answer was that we support > solutions not activities. > > Hiro: can we also support C code, Cobol, etc. The answer is yes, but > we need to support the environments. > > Ravi: we need to resolve the terms between the two groups prior to the > meeting. > > Hiro: there is a glossary document. Good suggestion (Dejan), that way > we can get consistent independently. It is linked off of > https://forge.gridforum.org/projects/ogsa-wg/docman/draft-ggf-ogsa-spe > c/ > en/23. > > Ravi: There is a space for provisioning in the EMS document, we try to > plug in the CDDLM as a virtual exercise. We try to effectively see if > CDDLM is compatible with OGSA. Chris, there are issues with containers. > > Chris: extending JSDL (extensions) with CDDLM or embedding the schemas. > Steve: it is batch -centric. Chris: it is not really. Ravi: keep > provisioning separate from JSDL. Dejan: summary, this is a good > proposal, but for simplicity reasons, we'll keep it out of our > immediate objectives for the time being. > > 2. Agenda for F2F (8/16, 1:30-5 in Sunnyvale): > > Action preparation for the meeting: sync up with glossary (Ravi: also > check up the glossary if we have any comments). > > A) (1:30-1:45) Introduction (learn more about EMS broader picture, in > the past wee overviewed CDDLM) > B) (1:45-3:15) Defining architectural relationship (This is really > related to the pictures we discussed above, Fig 5 in the OGSA arch & > Fig > 7 in CDDLM Foundation) > C) (3:15-3:45) Indebt overview of the CDDLM interfaces (if needed) > D) (3:45-4:15) Description and integration of ACS & CDDLM (Ravi: added > BES as well). > E) (4:15-4:45) Integration roadmap > F) (4:45-5:00) Next steps > > (Dejan: above times are approximate, they may change as needed, e.g. > if we cover some of the later items in earlier sessions) > > Ravi: is this similar to BES interfaces activities. Chris: this was > exactly how I was thinking wrt BES container. We can jump ahead to > conclusion that CDDLM also created container. it is definitely in the > context of BES cause we will run through the BES and we will > automatically se how it fits with BES. The same applies to RSF. It is > almost looking through a use case how do we do with something that is > given to job provisioning. > > Hiro: There is one more important component, ACS. Dejan/Steve, we came > to some agreement with ACS at GGF 14 and there is no need for changes > to CDDLM wrt ACS. Hiro: ACS had a F2F meeting in North Carolina. Mike > prepared a document describing relationship between CDDLM and ACS. Ravi: > we can add the ACS diagram to the figure. (as well as BES), this way > we can tackle multiple birds at the same time. Mike: ACS has taken > requirements wrt contents, there are some security considerations and > creation time issues that would require something other than URL, > maybe service invocation, we can go over these details at F2F. > > 3. Techniques and expectations: > > Dejan: informal approach, discussions, drawing pictures, agreeing on > interfaces, it is ok to have a few more open issues, but resolve them > at latest at GGF15. > > Ravi: can someone bring digital camera, take a snapshot and then send > it, put it on Webex. > > Hiro: expects architectural diagram and text describing components, so > that they can come into the next version of architectural document. > > Ravi: when is the time (Hiro: 1:30-5pm) > > >

2 1

Re: [gfs-wg] RNS critique (new revision available)
by Hiro Kishimoto 06 Aug '05

06 Aug '05

Takuya Ishibashi's email bounced -- Hiro Kishimoto To: gfs-wg(a)ggf.org Cc: ogsa-wg(a)ggf.org, ogsa-naming-wg(a)ggf.org Subject: Re: [gfs-wg] RNS critique (new revision available) From: Takuya Ishibashi <takuya(a)soum.co.jp> Hello Manuel and all, I have questions: * About QNames, do namespace in XML (and Prefix) for QNames have not to be defined strictly in this RNS document? Accoding to some examples in this document, namespace in XML for Required Entry Properties (in section 1.2.2.2.1) and Entry itself is "http://rns.ggf.org". And accoding to RNS WSDL 1.1, namespace in XML (for Inputs, Responses , Faults and ResourceProperties) is "http://rns.ws.ibm.com". * What is a XML datatype of Entry? According to EPRs and LogicalResolvers and Target (these are string) and some examples of <Entry>, I see that the XML datatype of Entry is string. (or any?) And, do the XML schema of Basic Namespace Components (in section 1.1) and Required Entry Properties (in section 1.2.2.2.1) have not to be provided in this RNS document? (as appendix?) Regards, Takuya Ishibashi

1 0

Re: [gfs-wg] RNS critique (new revision available)
by Manuel Pereira 05 Aug '05

05 Aug '05

Hello Osamu, > Now, all input to the service becomes a full path. Actually it always has been, however it has been presented in such a way that describes the use of an absolute path in all cases; this was done in an attempt to reduce complexity, potential confusion in a federated namespace, and present a consistent globally unique identifier for all namespace objects. (In Section > 1.3.3.1 "move", there still "the absolute path" remains in the second > table.) Right, I was originally thinking of allowing "moves" to occur between RNS providers, however this is simply not facilitated by the specification; a provider-to-provider interface would have to be described and that is out of scope for this revision. It is now changed, thanks! > This means the absolute path is only handled by a client > library not by a service itself. A service does not care about the > absolute path. Root directory is only known by a client. I do not > think there is any problem about this. If "handled" means that the client is the only party that maintains a comprehensive view of the overall federated namespace. The first sentence in section 2.2.1.1 clearly states that "all service operations [MUST] accommodate the use of absolute pathnames." This section goes on to describe how the referral message mechanism is employed to allow dynamic redirection to the appropriate RNS service providers if the path provided is not served by the current provider. > (In Section 1.4.6 "RNSJunctionFault", and in the first paragraph of > Section 2.2.1.1 "Referral Messages", "absolute" path should be "full" > path?) Please see the comments above, these two cases are correct and still valid. > Can we have a teleconference on Friday afternoon (3pm or 4pm) in > California time? This is a good idea and the time should work for me. Best regards, Manuel Pereira ------------------------------------------------------ IBM Almaden Research Center 1-408-927-1935 [Tieline: 457] mpereira(a)us.ibm.com ------------------------------------------------------

2 2

Re: [gfs-wg] RNS critique (new revision available)
by Manuel Pereira 05 Aug '05

05 Aug '05

Hello Takuya and all, > * About QNames, do namespace in XML (and Prefix) for QNames have not > to be defined strictly in this RNS document? This is a very good question. Since RNS has not yet been established as a standard, an official XML namespace has not been defined. This is orthogonal to the specification of the service however. Once the specification becomes a standard, then the appropriate XML namespace and WSDL references will be defined. > Accoding to some examples in this document, namespace in XML for > Required Entry Properties (in section 1.2.2.2.1) and Entry itself is > "http://rns.ggf.org". > > And accoding to RNS WSDL 1.1, namespace in XML (for Inputs, > Responses , Faults and ResourceProperties) is > "http://rns.ws.ibm.com". Actually, all of the examples that use any XML namespace use ?rns.ggf.org? since this particular specification draft is being submitted to GGF for standardization. The ?rns.ws.ibm.com? is only found in the WSDL, which is an actual working WSDL used to build the reference implementation, which is undergoing final development within IBM. The reality is that both of these may change. The official XML namespace may change once RNS becomes a standard, and the WSDL of the reference implementation will eventually match the defined standard. > * What is a XML datatype of Entry? > > According to EPRs and LogicalResolvers and Target (these are string) > and some examples of <Entry>, I see that the XML datatype of Entry > is string. (or any?) This is another good question. As outlined by the WSDL, <Entry> is not defined as a ComplexType. This simply means that Entry is an implicit type defined only in the semantic of the service. The <Entry> element (reflecting the Entry QName) is used to demarcate related property elements in a return message, but since the service is based on document style messaging, it does not define a specific ComplexType for Entry. When the property values associated with a particular entry are returned in a QueryResponse message, the property values are enclosed in an <Entry> envelope within the array of xsd:any elements. > And, do the XML schema of Basic Namespace Components (in section > 1.1) and Required Entry Properties (in section 1.2.2.2.1) have not > to be provided in this RNS document? (as appendix?) Only the QNames must be defined. Section 1.1 presents the ?Type? values possible for any namespace entry. These ?Type? values are literally identified by the QName defined. They are not objects but rather a property value or attribute of an object. For this reason, they are stored as strings and presented by the service in the form of a QName for consistent value comparison. Section 1.2.2.2.1 lists all entry property names. These ?names? are used in the composition of a query to denote which properties the requestor is interested in. Furthermore, they are used to envelope the property values returned in response messages. So these QNames are represented within the WSDL as entry properties only for the purpose of declaring the data types descriptively within the WSDL. I hope that I properly understood your questions, please let me know if I misunderstood you. Best regards, Manuel Pereira ------------------------------------------------------ IBM Almaden Research Center 1-408-927-1935 [Tieline: 457] mpereira(a)us.ibm.com ------------------------------------------------------

1 0