April 2011 - occi-wg - lists.ogf.org

Dasein Cloud
by Thijs Metsch 23 Jun '13

23 Jun '13

Hi, Just to inform you that there is a cool project going on called Dasein-Cloud. http://dasein-cloud.sourceforge.net/ On the website it reads: Dasein Cloud is an Open Source Java API made available by enStratus Networks LLC under the terms of the Apache License 2.0. This API specifies cloud-independent interfaces for enabling an application to access resources across multiple clouds. If you are interested have a look - and it might be great to implement OCCI in their someday! Cheers, -Thijs -- Thijs Metsch Senior Software Engineer Grid and Cloud Technology Platform Computing GmbH Europaring 60 D-40878 Ratingen http://www.platform.com http://www.nohuddleoffense.de/ - http://www.twitter.com/befreax

3 2

Validating the Cloud on, well, the Cloud
by alexander.papaspyrou＠tu-dortmund.de 14 Oct '11

14 Oct '11

Folks, in a moment of boredom, I rewrote Thijs' OCCI testing tool to run in Google App Engine (GAE). Please take a look at http://doyouspeakocci.appspot.com for the application (and, of course, try your service) and see https://github.com/lxndrp/doyouspeakocci/ for the source code of the application. The code is largely undocumented, but not very difficult to understand, if you are (like me) a fairly bad Python programmer and ever heard of Django Templates. If you find anything that you wish to improve, feel free to do so. A few rules of thumb for this: 1. Although the code is released under LGPL2, I consider it "AdviseWare": when making an improvement, let me know where I was too stupid to get it right in the first place. 2. Remember that GAE has certain restrictions in what can be done and how you can do it. See http://code.google.com/intl/de-DE/appengine/docs/ for details. 3. Before making a change, create a ticket in the GitHub tracker. 4. During making a change, remember to add some documentation (for new classes and defs) or at least comments (close to your change). 5. After making a change, enjoy the fuzzy feeling of having contributed to the community. Even if you are not a Python developer, you can help out: the pages served by the application look, ahem, bad. So if you happen to be an HTML/CSS expert, your help is more than appreciated. Ultimately, the testing tool should look similar to the web pages at http://occi-wg.org (ask Andy whether you can feel free to steal stuff from there for the GAE app!). Hoping you will enjoy the tool, my mind demands some rest now and advises me to be off for a few beers. Best, Alexander

7 11

Where is the scheme ?
by Gary Mazz 04 Oct '11

04 Oct '11

Hi There is no scheme @ http://schemas.ogf.org/occi/ cheers, gary

7 15

occi4java implementation experiences document
by Sebastian Heckmann 27 Apr '11

27 Apr '11

Hi everyone, our occi4java is now on github (https://github.com/occi4java/occi4java) under the LGPL license and we are ready to write an implementation experience guide, like: http://ogf.org/documents/GFD.176.pdf Are there any ideas about the content? It would be very helpful for us. Thanks in advance Sebastian Heckmann Sebastian Laag

4 3

Cloud Plugfest summary
by florian.feldhaus＠tu-dortmund.de 25 Apr '11

25 Apr '11

Hi everyone, during the Cloud Plugfest a few issues came up which I'd like to summarize here. Quite a few are related to the OCCI verification suite. General discussion: - Location: * Currently there are 2 types of Locations. Once the location of Kinds, Mixins and Actions and on the other hand the Location of resources. The latter ones are returned to the OCCI client as absolute URI. This probably comes from RFC 2616 Sec. 9.5 where HTTP POST is defined. Interestingly, the creation of a entity should return a 201 (created) status code, whereas in the OCCI HTTP rendering document always 200 (OK) is returned. In the RFC for 201 an absolute URI for Location must be returned. This behaviour should be adapted in the OCCI specs. This applies also to PUT when a mixin is added or a resource is created. RFC 2616 is even more specific here and says "If a new resource is created, the origin server MUST inform the user agent via the 201 (Created) response" * The Location header field only is used for POST during resource creation. In all other cases, X-OCCI-Location is returned. At first this was irritating for me, but after reading RFC 2616 I understood, that Location can only return one entity and is recommended for POST during creation of an entity. It may help to clarify this in the OCCI specs. * As the entity Location returned by the OCCI implementation should be absolute, it needs to contain the port of the OCCI implementation if the port is not default (e.g. 80). Mixing of HTTP and HTTPS is thus not a good idea, as the namespaces would differ. On the other hand, one could require, that the client keeps track of the ports and knows if a non standard port is used. Currently this is an issue for the OCCI verification suite as it uses the URI returned by POST to do a GET on a newly created entity. This should be clarified in a best practice guide. - Namespace: * It might be a good idea to introduce a namespace discovery mechanism. Maybe all categories (e.g. kinds/mixins) could have an additional namespace attribute (e.g. containing "vms" for the compute kind or "disks" for the storage kind). This becomes more complex if patterns like /vms/$username are used. What do you think? OCCI verification suite (see attached patch for all changes applied to occi_test.py) - all tests * class= is missing for scheme in all Category headers. The definition of Category (3.5.1 of HTTP Rendering document) requires class="action" | "mixin" | "kind". A server could ommit checking the class parameter, but a client shouldn't rely on the server being sloppy. * check recent changes to attributes e.g. "occi.core.source" instead of "source" (Link). A few last minute changes were made to the Core and Infrastructure documents with regards to attribute names. These need to be adopted. - test_infrastructure_model_for_completness * Categories of Actions shouldn't be part of GET /-/ (see OCCI HTTP Rendering 3.4.1) * Mixins shouldn't be related to links (see OCCI Core 4.4.3 Table 3) - so the ipnetworking mixing shouldn't have rel='http://schemas.ogf.org/occi/core#link' and instead rel='' - test_filter * heads['Content-Type'] = 'text/occi' is missing, this is a problem if the test is run standalone * delete -> location needs to include port? See comment above - test_links * source and target must be replaced with occi.core.source and occi.core.target for Link: link['X-OCCI-Attribute'] = 'occi.core.source=' + compute_loc + ',occi.core.target=' + network_loc - test_actions * after action has been triggered, it should be checked if the status of the compute entity has been changed - test_create_kinds * must be occi.compute.hostname="foo" (with quotes) instead of occi.compute.hostname=foo (see OCCI HTTP rendering 3.5.4) OCCI HTTP Rendering document - Section 3.4.4.: "Triggering an Action on a resource instance": Category for Action must be "Category: stop; scheme="[...]"; class="action";" instead of "Category: compute; scheme="[...]"; class="action";" (see OCCI Infrastructure 3.1 Table 3) - Section 3.4.1.: Adding a Mixin definition: Must be rel="http:/example.com/occi/something_else#mixin;" with quotes OCCI Core document - 4.4.3 TYPO: "Mixin might by applicable" ->" Mixin might be applicable" Cheers, Florian

2 1

Cloud Plug Fest Remote Meeting Details - please connect and test your OCCI implementations!
by Alan Sill 20 Apr '11

20 Apr '11

OCCI-WG folks, (this means you!) Can any and all able to connect and contribute to the plug-fest as below please do so? The event runs only today and tomorrow, AFAIK, so this is our chance to shine! I will not drive from here to Colorado Springs for one day, all that I have left after recovering (feeble excuse but true), and I think what we need now are heavy-hitting contributions from real implementations, so please do call in as below and contribute! Alan On Apr 20, 2011, at 3:17 PM, Mark Carlson wrote: > Can you join the chat room we set up: > http://webconf.soaphub.org/conf/room/CloudPlug > > Also we will have a sync up call today at 4:30pm MT. > > The OCCI guys are struggling... > > -- mark > > On 4/17/11 9:59 PM, Sill, Alan wrote: >> FYI - see call-in details below. >> >> I picked up some kind of bug and will likely not make it there in person, at least for the first part of the meeting, as I had hoped to do. Antibiotics are kicking in, though, so I hope to join remotely at minimum for at least part of it. Look forward to some of you joining in also. >> >> Alan >> >> Begin forwarded message: >> >> From: Mark Carlson < >> mark.carlson(a)oracle.com<mailto:mark.carlson@oracle.com> >> > >> Date: April 17, 2011 12:17:16 PM CDT >> To: " >> cloudplugfest(a)snia.org<mailto:cloudplugfest@snia.org>" <cloudplugfest(a)snia.org<mailto:cloudplugfest@snia.org> >> > >> Subject: Cloud Plug Fest Remote Meeting Details >> >> >> The Conference call info is the same all days: >> ------------------------------------------------------- >> Teleconference information >> ------------------------------------------------------- >> Call-in toll-free number: 1-(866) 682-4770 (US) >> Call-in number: 1-(408) 774-4073 (US) >> Show global numbers: >> <https://www2.audiocontrols.net/WebExAdaptor/globalnumber?participantcode=12…> https://www2.audiocontrols.net/WebExAdaptor/globalnumber?participantcode=12… >> >> >> Conference Code: 1212488# >> >> Security: 1212# >> >> I will start the webex during the specific times below, and can always start one on demand if needed. If the room is not on the above concall number, please call me at: 303-720-6139 and I will start it up. >> >> Topic: Cloud Plug Fest Tuesday Morning >> Date: Tuesday, April 19, 2011 >> Time: 8:30 am, Mountain Daylight Time (Denver, GMT-06:00) >> Meeting Number: 595 238 426 >> Meeting Password: cloud2011 >> >> ------------------------------------------------------- >> To start or join the online meeting >> ------------------------------------------------------- >> Go to >> <https://oracleweb.webex.com/oracleweb/j.php?ED=158097672&UID=494227547&PW=N…> https://oracleweb.webex.com/oracleweb/j.php?ED=158097672&UID=494227547&PW=N… >> >> >> Topic: Cloud Plug Fest Tuesday Sync Up >> Date: Tuesday, April 19, 2011 >> Time: 4:30 pm, Mountain Daylight Time (Denver, GMT-06:00) >> Meeting Number: 596 468 228 >> Meeting Password: cloud2011 >> >> ------------------------------------------------------- >> To start or join the online meeting >> ------------------------------------------------------- >> Go to >> <https://oracleweb.webex.com/oracleweb/j.php?ED=158097682&UID=494227547&PW=N…> https://oracleweb.webex.com/oracleweb/j.php?ED=158097682&UID=494227547&PW=N… >> >> >> Topic: Cloud Plug Fest Wed Sync Up >> Date: Wednesday, April 20, 2011 >> Time: 4:30 pm, Mountain Daylight Time (Denver, GMT-06:00) >> Meeting Number: 599 145 796 >> Meeting Password: cloud2011 >> >> ------------------------------------------------------- >> To start or join the online meeting >> ------------------------------------------------------- >> Go to >> <https://oracleweb.webex.com/oracleweb/j.php?ED=158098317&UID=494227547&PW=N…> https://oracleweb.webex.com/oracleweb/j.php?ED=158098317&UID=494227547&PW=N… >> >> >> Topic: Cloud Plug Fest Thursday Matrix >> Date: Thursday, April 21, 2011 >> Time: 10:00 am, Mountain Daylight Time (Denver, GMT-06:00) >> Meeting Number: 590 705 049 >> Meeting Password: cloud2011 >> >> ------------------------------------------------------- >> To start or join the online meeting >> ------------------------------------------------------- >> Go to >> <https://oracleweb.webex.com/oracleweb/j.php?ED=158098322&UID=494227547&PW=N…> https://oracleweb.webex.com/oracleweb/j.php?ED=158098322&UID=494227547&PW=N… >> >> >> Topic: Cloud Plug Fest Thur Sync Up >> Date: Thursday, April 21, 2011 >> Time: 4:30 pm, Mountain Daylight Time (Denver, GMT-06:00) >> Meeting Number: 599 526 436 >> Meeting Password: cloud2011 >> >> ------------------------------------------------------- >> To start or join the online meeting >> ------------------------------------------------------- >> Go to >> <https://oracleweb.webex.com/oracleweb/j.php?ED=158098327&UID=494227547&PW=N…> https://oracleweb.webex.com/oracleweb/j.php?ED=158098327&UID=494227547&PW=N… >> >> >> >

1 1

Fwd: Microsoft's contribution to DMTF's CMWG
by Alan Sill 20 Apr '11

20 Apr '11

FYI - comments welcome. Alan Begin forwarded message: > From: Babak Jahromi <babakj(a)microsoft.com> > Date: April 10, 2011 1:30:16 AM EDT > To: "Sill, Alan" <alan.sill(a)ttu.edu>, "Craig.A.Lee(a)aero.org" <Craig.A.Lee(a)aero.org>, "tmetsch(a)platform.com" <tmetsch(a)platform.com> > Cc: Colleen Evans <coevans(a)microsoft.com> > Subject: Microsoft's contribution to DMTF's CMWG > > Gentlemen, > > Last week Microsoft made the attached contribution to CMWG for cloud management model. It is based on using DMTF’s CIM on top of OData. > > To learn more about the OData effort: http://odata.org. You can also join a public mailing list to participate in discussions on the evolution of OData. The mailing list info is at: http://www.odata.org/mailing-list > > If you have any question, please feel free to ask Colleen or me. Colleen actually represents Microsoft in CMWG. > > Thanks, > Babak

2 1

No concall today - Switch to bi-weekly concalls?
by Thijs Metsch 20 Apr '11

20 Apr '11

Hi all, There will not be a conference call today. For next week we might have something to share with you all with regards to Monitoring/SLA/negotiation features based on OCCI. I would like to ask all of you if it would be okay to switch to a bi-weekly schedule for our calls? If no objections are heard till next week - we'll start doing our calls every other week. Thanks, -Thijs

1 0

OGF - DMTF work register update: Draft of 18 April 2011
by Alan Sill 19 Apr '11

19 Apr '11

Attached is a draft of the OGF - DMTF work register updated to reflect input from the OCCI OGF work group and to eliminate work completed since the last update, or otherwise inapplicable in light of current activities. At the request of the OCCI group, I have also eliminated a large set of general text that seemed only intended to make introductory statements, and trimmed much of the rest. In this spirit, the "Alliance Benefits" section could also be eliminated, but for now I have left it in place. I'd like to ask for your review on the content and specifics of the attached document. Mark would like a version to send to the DMTF board no later than Tuesday, so I hope that our work to this point has brought it to the state that you can make very specific comments. I would like to note that this should be a mutual document, in other words reflect work to be done within both organizations, not just be a one-sided document in which OGF (for example) spells out its own work without input from DMTF. Nonetheless I hope we are close to the start of a substantial agreement that will allow us to carry out the next phase of the work. Thanks, Alan Alan Sill, Ph.D Senior Scientist, High Performance Computing Center Adjunct Professor of Physics, TTU VP of Standards, Open Grid Forum

5 4

Quick note on next steps
by Thijs Metsch 18 Apr '11

18 Apr '11

Dear all, Beside all the cool extensions we're gonna be working on soon we need to write something like this sometime: http://ogf.org/documents/GFD.176.pdf This experience document is taken from the SAGA group, and is a good example - Experience documents are part of the OGF process BTW... Just wanted to throw this in based on current ongoing discussions... Cheers, -Thijs

2 1