March 2012 - occi-wg - lists.ogf.org

Happy St. Patrick's Day!
by Edmonds, AndrewX 16 Mar '12

16 Mar '12

http://occi-wg.org/2012/03/16/happy-st-patricks-day/ PS: Lots of new posts up on the blog and more to come! ------------------------------------------------------------- Intel Ireland Limited (Branch) Collinstown Industrial Park, Leixlip, County Kildare, Ireland Registered Number: E902934 This e-mail and any attachments may contain confidential material for the sole use of the intended recipient(s). Any review or distribution by others is strictly prohibited. If you are not the intended recipient, please contact the sender and delete all copies.

1 0

Updates to OCCI core
by Feldhaus, Florian 14 Mar '12

14 Mar '12

Here are some ideas for extending the description of attributes. To more clearly specify attributes, a new new class 'attribute' should be added to OCCI Core. For the class diagram this would require to remove '+ attributes :Set<String>' from category and instead connect it with the attribute class in a many to one dependency. Ideas for the attribute class: - the attribute class should consist of the attributes: mutable, required, type, default - the naming of attributes of the OCCI specification MUST start with occi followed by term and the attribute name (e.g. occi.compute.cores) - the naming of attributes of user defined mixins MUST start with the reversed domain name of the mixin scheme followed by the term and the attribute name (e.g. com.example.my_mixin.my_attribute) - the attribute type should contain restrictions on the attribute like for example: * string: choices, min_length, max_length * number: bounds (one or more lower and upper bounds) - there might be more detailed attribute types such as URL or URI, float, integer, ... --Florian ------------------------------------------------------------------------------- GWDG - Gesellschaft für wissenschaftliche Datenverarbeitung mbH Göttingen Am Fassberg 11, 37077 Göttingen Fon: 0551 39-20364 Fax: 0551 201-2150 E-Mail: florian.feldhaus(a)gwdg.de WWW: www.gwdg.de ----------------------------------------------------------------------------------- Geschäftsführer: Prof. Dr. Ramin Yahyapour Aufsichtsratsvorsitzender: Prof. Dr. Christian Griesinger Sitz der Gesellschaft: Göttingen Registergericht: Göttingen Handelsregister-Nr. B 598 -----------------------------------------------------------------------------------

4 4

OGF34 OCCI Sessions
by Edmonds, AndrewX 14 Mar '12

14 Mar '12

Hi all, We're having significant networking issues here in Oxford so we cannot host a skype confcall. Apologies. Andy ------------------------------------------------------------- Intel Ireland Limited (Branch) Collinstown Industrial Park, Leixlip, County Kildare, Ireland Registered Number: E902934 This e-mail and any attachments may contain confidential material for the sole use of the intended recipient(s). Any review or distribution by others is strictly prohibited. If you are not the intended recipient, please contact the sender and delete all copies.

1 0

OCCI Extensions, NIST and others
by Edmonds, AndrewX 12 Mar '12

12 Mar '12

Hey all, At the SNIA Cloud Plugfest Florian and I went through a number of required extensions that are needed for closer compliance with the NIST cloud computing VM use cases. I've documented them here [1]. Also included in the document is the beginnings of OVF capabilities and updates to CDMI interoperability. Comments etc gladly welcome. Also we can talk about these and others at the Weds. sessions at OGF34. Andy [1] https://docs.google.com/document/d/1XnUssN8nYUAwvuuI-3WGmbtIR-O33RS0gF9ZFed… ------------------------------------------------------------- Intel Ireland Limited (Branch) Collinstown Industrial Park, Leixlip, County Kildare, Ireland Registered Number: E902934 This e-mail and any attachments may contain confidential material for the sole use of the intended recipient(s). Any review or distribution by others is strictly prohibited. If you are not the intended recipient, please contact the sender and delete all copies.

1 0

Summary of the JSON rendering discussion
by Feldhaus, Florian 12 Mar '12

12 Mar '12

Following I try to summarize the recent discussion of the JSON rendering, please feel free to add, extend or correct it: - different content types * different content types make it easier to identify the content of the message * content type negotiation may get more complicated * for GET requests, the client may know what type of content he's requesting (even though he may use different content types in Accept and doesn't know which one the server responds with) * for POST requests, the server doesn't know what content type he gets - if just application/json is used as media type (instead of e.g. application/occi+json or more detailed content types), the client doesn't know which semantics the content have - CDMI uses specific media types to help client and server to know the semantics of the content without first interpreting the content - how should the location of the entities be named? location, href, uri? - there should be a recommendation that the JSON format should be rendered compact (e.g. without unnecessary whitespaces) in production use - how should the category identifier be called? rel, type? rel is based on the HTTP way of doing things - make the JSON format as easy to use and as compact as possible - there was a security issue with having JSON arrays at the top level. This issue is solved as of ECMAScript5 [1]. Using just an array with all categories or entities would be a compact rendering. Questions are: -- Is ECMAScript5 already widely adopted? -- when exactly does this security issue appear? Can it be exploited for HTTPS connections where the OCCI endpoint has been clearly identified by a valid certificate? - should there be a JSON rendering of Locations or is it sufficient to have text/uri-list? Could a JavaScript implementation easily parse text/uri-list if the client intends to just get the locations of some object (maybe together with a filter)? - rendering of entities: -- is rel necessary for actions? -- which fields are really necessary for links inside of entities? --- smallest version would be just the link location --- title would be very handy for displaying in user interfaces --- title, target uri/href, rel, link href/uri, link rel and attributes would make it easier for the client to directly use links, but may introduce a lot overhead for huge number of links -- attributes should be rendered according to their hierarchy (e.g. as { 'occi': { 'core': { 'id': '123' } } } ). This makes using attributes very simple for clients, especially those written in JavaScript. -- all attributes must be unique regardless how they are rendered. This should be taken care of in OCCI Core and may be achieved by using reversed domain names for user/provider defined mixins. -- depending on the impact of the security issue with Top Level arrays, entities can be rendered either compact with only an array or a little more verbose by using a JSON object/hash with key 'entities' and the array of entities as value -- should single entities be rendered differently from multiple entities? e.g. should a single entity be directly rendered as JSON object/hash with kind, mixins, … or should it be in the same format as multiple entities e.g. inside an array? - rendering of categories: -- should attributes be rendered as single string or as hierarchy of hashes (e.g. "occi.core.id" : ... vs. "occi": { "core: { "id": … } } }? -- type should be extended with the restrictions defined in OCCI Core for the attribute type. This can be e.g. for string: choices / min_length, max_length and for numbers: bounds Example of proposed compact rendering (save whitespaces) of OCCI entities by Jamie, Florian and others: http://pastebin.com/aun0Ha2D Combination of above rendering with Ralfs Collection format (replacing 'collection' to 'entity' because I don't like using the name collections when its an array of entities ;-) ): http://pastebin.com/bffFtL19 Example of proposed category rendering by Ralf, extended by Jamie, Florian and others: http://pastebin.com/txhnt7uX [1] http://flask.pocoo.org/docs/security/#json-security --Florian ------------------------------------------------------------------------------- GWDG - Gesellschaft für wissenschaftliche Datenverarbeitung mbH Göttingen Am Fassberg 11, 37077 Göttingen Fon: 0551 39-20364 Fax: 0551 201-2150 E-Mail: florian.feldhaus(a)gwdg.de WWW: www.gwdg.de ----------------------------------------------------------------------------------- Geschäftsführer: Prof. Dr. Ramin Yahyapour Aufsichtsratsvorsitzender: Prof. Dr. Christian Griesinger Sitz der Gesellschaft: Göttingen Registergericht: Göttingen Handelsregister-Nr. B 598 -----------------------------------------------------------------------------------

1 0

Updates to OCCI specification
by Feldhaus, Florian 12 Mar '12

12 Mar '12

Hi, picking up the recent discussion on JSON, I'd like to summarize the current ideas as far as I understood them from the recent discussion. First, there was the proposal to put the changes in the right documents as errata. I very much second this, even if we may face problems pushing it through OGF as errata. Thus I'd like to split the ideas into the corresponding parts of the OCCI and send separate mails for them. It would be nice if we could keep the threads separate. These are currently ideas and I might have missed some of them. Please add all ideas I missed and try to focus on getting everything ready for integration into the specification as soon as possible. There are several implementations ready to pick up especially the JSON rendering. Also it would be great if we could get OCCI from proposed recommendation to full recommendation as soon as possible. Cheers, Florian ------------------------------------------------------------------------------- GWDG - Gesellschaft für wissenschaftliche Datenverarbeitung mbH Göttingen Am Fassberg 11, 37077 Göttingen Fon: 0551 39-20364 Fax: 0551 201-2150 E-Mail: florian.feldhaus(a)gwdg.de WWW: www.gwdg.de ----------------------------------------------------------------------------------- Geschäftsführer: Prof. Dr. Ramin Yahyapour Aufsichtsratsvorsitzender: Prof. Dr. Christian Griesinger Sitz der Gesellschaft: Göttingen Registergericht: Göttingen Handelsregister-Nr. B 598 -----------------------------------------------------------------------------------

3 2

Updates to OCCI HTTP Rendering
by Feldhaus, Florian 11 Mar '12

11 Mar '12

Based on the proposal from Ralfs JSON Rendering document the following additions should be added to the HTTP Rendering: - filtering and pagination for GET requests (e.g. attribute-filter, category-filter, pagination-marker, pagination-limit) An idea of Ilja Livenson was, to add a 'depth' query string as well, to allow for the retrieval of a restricted depth in the location hierarchy. --Florian ------------------------------------------------------------------------------- GWDG - Gesellschaft für wissenschaftliche Datenverarbeitung mbH Göttingen Am Fassberg 11, 37077 Göttingen Fon: 0551 39-20364 Fax: 0551 201-2150 E-Mail: florian.feldhaus(a)gwdg.de WWW: www.gwdg.de ----------------------------------------------------------------------------------- Geschäftsführer: Prof. Dr. Ramin Yahyapour Aufsichtsratsvorsitzender: Prof. Dr. Christian Griesinger Sitz der Gesellschaft: Göttingen Registergericht: Göttingen Handelsregister-Nr. B 598 -----------------------------------------------------------------------------------

1 0

Re: [occi-wg] NIST Cloud Metrics team initial survey of vendor interfaces and SDO-supplied cloud standards-specified capabiltiies
by Alan Sill 09 Mar '12

09 Mar '12

I forgot to mention the oddity of this twiki page - the attachments only show up if you click the tiny reveal triangle next to the word "attachments" Here is a URL to this file: http://collaborate.nist.gov/twiki-cloud-computing/pub/CloudComputing/RATax_… or http://tinyurl.com/7lwpzet Alan On Mar 9, 2012, at 4:18 PM, Mark Carlson wrote: > Alan, > > Do you have an actual URL to > > Cloud-_Cloud_Standards_-_Cloud_API_Comparison_jeffs_edits.xls > > ? > > I can't seem to find it on the web site. > > -- mark > > > On 3/9/12 9:00 AM, Alan Sill wrote: >> Dear OCCI group members, >> >> FYI, a survey of SDO offerings and vendor capabilities has been started by a member of the NIST Cloud Metrics standards analysis sub-team. You might want to look at the documents at the following link, especially the one labeled >> >> Cloud-_Cloud_Standards_-_Cloud_API_Comparison_jeffs_edits.xls >> >> for accuracy and completeness, and please comment. (I have so far not made any input into this document, and would like to invite your review and opinions please). >> >> Topic: >> RATax_CloudMetrics_Meeting_03082012< CloudComputing< TWiki >> NIST Cloud Metrics team >> >> Link: >> http://collaborate.nist.gov/twiki-cloud-computing/bin/view/CloudComputing/R… >> >> these things have a way of migrating around, so I would like us to be able to correct any mis-impressions, omissions or other errors ASAP please. >> >> Hope to see you all soon at OGF 34! >> >> Thanks, >> Alan >> >> _______________________________________________ >> occi-wg mailing list >> occi-wg(a)ogf.org >> https://www.ogf.org/mailman/listinfo/occi-wg

1 0

NIST Cloud Metrics team initial survey of vendor interfaces and SDO-supplied cloud standards-specified capabiltiies
by Alan Sill 09 Mar '12

09 Mar '12

Dear OCCI group members, FYI, a survey of SDO offerings and vendor capabilities has been started by a member of the NIST Cloud Metrics standards analysis sub-team. You might want to look at the documents at the following link, especially the one labeled Cloud-_Cloud_Standards_-_Cloud_API_Comparison_jeffs_edits.xls for accuracy and completeness, and please comment. (I have so far not made any input into this document, and would like to invite your review and opinions please). Topic: RATax_CloudMetrics_Meeting_03082012 < CloudComputing < TWiki NIST Cloud Metrics team Link: http://collaborate.nist.gov/twiki-cloud-computing/bin/view/CloudComputing/R… these things have a way of migrating around, so I would like us to be able to correct any mis-impressions, omissions or other errors ASAP please. Hope to see you all soon at OGF 34! Thanks, Alan

1 0

Session invitation for next week
by alexander.papaspyrou＠tu-dortmund.de 05 Mar '12

05 Mar '12

Folks, as most of you might be aware, OGF is coming closer: next week, the event will take place in Oxford, UK. DCIfed-WG has a session registered for Wednesday, March 14 at 09:00h BST. Currently, DCIfed embraces three topics: 1. The original DCIfed cross-domain interop use cases (as expressed in the charter) This has been largely done by the DGSI project in Germany, which already has produced a number of documents on how to use available technologies from within OGF to achieve this. I consider this aspect mostly paperwork, bringing the docs into the OGF recommendation format. I would really appreciate if people (especially those with DGSI background) could volunteer to help with the paperwork. We need to discuss with our AD whether the docs should be GFD.Is or GFD.Rs, but this can be done in the session. Besides this, I am not planning to give a presentation on this (I have done so several times before), unless there is explicit interest for it -- please express this on the list. 2. The EGI Federated Cloud Task Force interop use cases (as proposed in [1]): This is the more "cloudy" part of DCIfed work, which emerged from David Wallom's IaaSCIP-BoF in Taipei last year. Currently, Matteo Turilli from OeRC is leading this task in FedCloud-TF. They have made some progress in terms of implementation, and I think time has come to start working on a profile document in the OGF document process to produce a P-REC as well. I would therefore like to invite the FedCloud-TF people to save the date and join the session so that we can make some progress. Also, Matteo, it would be good if you (or someone else from the group) could give a talk on FedCloud-TF and the state of the union. I think David already has a deck on the general ideas (he showed one at the last OGF in Lyon), and an update on what's on the wiki would make this a nice intro. Please let me know how much time you would like to have for this. 3. The Cloud Interoperability BoF (submitted by Florian Feldhaus) You might be aware that there is another Cloud-related BoF on Tuesday, led by Florian Feldhaus. Although the charter of DCIfed had been written in a very broad sense, the developments over time have shown that, in fact, the main work will move into the Cloud direction anyway (mostly due to the FedCloud-TF topic). I would therefore like to invite Florian to the session to hear about the results from the BoF and to discuss how and to what extent this relates to the FedCloud-TF work. In that context, we might as well consider rechartering of the WG towards this area of interest. Florian, please let me know if you are available, and whether you would be willing to give an update on the group. Looking forward to see you in Oxford! Cheers, Alexander -- Alexander Papaspyrou alexander.papaspyrou(a)tu-dortmund.de

1 0