Regarding the SLA@SOI implementation of OCCI the following videos show how we use OCCI in a more PaaS style. OCCI powers the infrastructure service that provides virtual machines that are guaranteed by automated SLAs. http://www.youtube.com/watch?v=1oGwGBqbk4Y http://www.youtube.com/watch?v=NZhQ4sl1CTg The OCCI implementation that backs the demo is one that is based on OCCI 1.0 with additional SLA@SOI functionality added. Currently the SLA@SOI implementation is being upgraded to OCCI v1.1. In terms of OpenNebula and OCCI, Florian Feldhaus and his team from UDO having been updating OpenNebula to provide full OCCI v1.1 compliance and gave us a good demonstration of the implementation at yesterday's OCCI confcall (thanks Florian!). That code will soon be made public. HTH, Andy PS: The SLA@SOI implementation and the SSF implementation are separate ones. -----Original Message----- From: occi-wg-bounces@ogf.org [mailto:occi-wg-bounces@ogf.org] On Behalf Of Thijs Metsch Sent: Thursday, March 31, 2011 10:28 AM To: Sill, Alan Cc: Tong, Jin; Sokol, Annie W.; Liu, Fang; Badger, Mark Lee; Hogan, Michael D.; occi-wg@ogf.org Subject: Re: [occi-wg] Current drafts and comments on "pre-9th" draft for Standards Roadmap document; meeting schedule Hi,
And the implementation of OCCI in OpenStack and Eucalyptus are still under way and not released, correct?
The OpenStack implementation is still under development - I rarely have time to work on it sadly enough. I added David Wallom to the CC list, he might be able to give an update about OCCI in Eucalyptus.
* Is there any implementations of OCCI to offer PaaS service management?
Well since OCCI is basically a boundary protocol (See http://occi-wg.org/about/) offerings supporting IaaS, PaaS and SaaS can be easily created using OCCI. Most implementation support this. E.g. the SSF (http://pyssf.sf.net) has an example on a keyvalue store or job submission. To my last knowledge Ralf Nyren's implementation will support Redis soon.
* In OCCI process, is there ever any consideration of delivering some reference implementation?
Hi Jin,
Thanks for your questions. My not-very-well-educated answers (I'm more on top of the document process than the actual implementations themselves) are embedded inline below; beyond this I invite the OCCI-WG members to comment
Please see this web page: http://occi-wg.org/community/implementations/ As Alan stated there are several implementations around already. The currently two most up to date can be tested here: http://www.nyren.net/api/ and http://fjjutraa.joyent.us:8888/ (Both python implementation installable through easy_install occi|pyssf). At least 4 implementations are currently finalizing there last changes and should have interfaces up soon! So there is no real reference implementation but rather a pool to choose from. We have some compliancy test code here - which is currently OCCI bound - but could be reused: http://occi-wg.org/2011/01/18/occi-compliance-testing-tool/ Also handy in case of testing compliancy is the ANTLR grammar Andy Edmonds created for the SLA@SOI project. That can be found here: https://github.com/dizz/occi-grammar Hope this helps, -Thijs On Mar 31, 2011, at 4:43 AM, Sill, Alan wrote: themselves.
On Mar 30, 2011, at 7:07 PM, "Tong, Jin" <jin.tong@nist.gov> wrote:
Alan, thanks for sharing your comments. I have a few questions about OCCI
* My understanding is OpenNebula has OCCI implemented, as functional
equivalent to EC2 interface for VM management, correct?
Yes. We have definite commitments from the primary project leader for OpenNebula for full implementation of all current OCCI features - their
adoption and implementation status after going through the presentation materials linked from your email: present product includes support for an older version - and in his words, to "build an ecosystem" of OpenNebula functionality around OCCI.
And the implementation of OCCI in OpenStack and Eucalyptus are still
under way and not released, correct?
I have seem a screenshot of a login window from an OpenStack
implementation claiming to offer OCCI services. Beyond this and some email confirming it is on the roadmap for the bexar release, I don't know any details on progress in OpenStack. The Eucalyptus implementation is being paid for by a UK-funded project that includes some other significant enhancements for use of Eucalyptus in science infrastructure projects; my impression is that it is underway but still perhaps a few months off.
* Is there any implementations of OCCI to offer PaaS service management?
Of the implementations I have taken the time to look at, the SSF one by
SLA@SOI seems to me to be the closest to offering connections to services at the PaaS and actual application level. Here I really don't want to stray into territory where I'm not an expert, though, and invite the developer community to respond. Note that several products that have implemented OCCI or are committed to doing so, like OpenNebula for example, could themselves be considered as PaaS products.
* In OCCI process, is there ever any consideration of delivering some
reference implementation?
OGF as an SDO itself never delivers software, but concentrates on the
process of developing communities that develop work products (documents), and encouraging implementations to be developed, then documented themselves as part of the experience process and refinements if ant to the original proposed recommendation before promotion to a full OGF Recommendation. The latter category requires at least two independent implementations and an extended period of practical use (at minimum, 6 months) plus an experience document comparing implementation experiences in the field before promotion to a full REC. As we already have so many implementations to compare before even full publication of the OCCI P-REC specs, I am not worried at this point about having enough material for the subsequent steps!
So while we do not have the official process in OGF of providing official
reference implementations, I have asked the group to help provide some test code that can be run against a demo instance of an OCCI-interfaced service that could be used, for example, in SAJACC. Among the options to get something going quickly for one such implementation (with quite a general package name, but just one option among many) would be to type "easy_install occi" to get Ralf Nyren's pypi OCCI package.
https://github.com/nyren/occi-py#readme
The availability of the libvirt implementation by TU-Dortmund and
popularity of libvirt as an underlying layer for IaaS products should make other implementations easy to produce for libvirt-based products.
At this point, however, I want to get out of the way as quickly as
possible and invite any OCCI group members to comment, especially if they have answers to your questions or want to correct any mistakes I have made!
Take care and best wishes, Alan
Thanks! --Jin ________________________________________ From: cc_standards@nist.gov [cc_standards@nist.gov] On Behalf Of Alan
Sill [Alan.Sill@ttu.edu]
Sent: Wednesday, March 30, 2011 12:55 PM To: CC_STANDARDS Subject: Current drafts and comments on "pre-9th" draft for Standards Roadmap document; meeting schedule
Hi Mike and Annie,
First of all, thank you both very much for your great leadership and coordination in the NIST Standards Roadmap working group. Thanks also to the group for their great work so far.
Due to travel, I have not had the opportunity to participate in recent meetings, but see that there has been great progress on the document. The text and graphical descriptions have gotten much better, and I see that there is movement in filling out some information on the available standards.
As you know, OGF is close to final release of the Core and Infrastructure components of the OCCI specification set, and I see that it has already been incorporated into the tables as what I expect it will be by the time of the NIST meeting, as an Approved Standard. (This is correct; OGF terminology for a newly released standard is "Proposed Recommendation" (P-REC), two implementations plus documented feedback from extensive actual use in the community for an extended period are required to promote a GFD document to a full REC.)
I have asked the OCCI working group to review your current draft and to provide feedback on the comment template form for any alterations they would like to suggest. At this point I have only asked them for suggested changes in the interest of accuracy. As you know, we have had a tremendous uptake of OCCI by several open source projects and now two commercial ones; this is probably due to its simplicity as an extended set of features to http headers, and its consequent explicit platform neutrality and language independence leading to a great ease of implementation. (My last count was 18 OCCI implementations either completed or in advanced states of work, and growing fast.)
We certainly want to document this progress adequately and hope to provide a proper set of comments to the draft roadmap soon, with earnest hope for their incorporation.
I note that there is not a meeting listed in the schedule for the Roadmap group for this week, which would normally be tomorrow (March 31). In light of the possible availability of the comments mentioned above, and to give the group another chance to look over the draft document, can I suggest that we meet, either as a group or just as a sub-group to look a these comments, at the usual Standards Roadmap meeting time of 1 pm Eastern tomorrow to discuss these?
Let me know.
Finally, please note that OGF has more standards of interest to clouds and other large-scale distributed computing infrastructures than just OCCI. Since this type of work is the focus of our organization as an SDO, we have many other specifications that have good uptake in industry; in particular, the secured transport-level data transfer specifications SRM and GridFTP, the WS-* based Basic Execution Services family of specs including those in the OGSA series, and our WS-Agreement and WS-AgreementNegotiation series oriented toward automated service level agreement negotiation, license management, etc. are probably also of interest. The last one mentioned (WS- Negotiation for short) is in final public comment until May 15, 2011 at http://ogf.org/gf/docs/?public_comment or directly at http://ogf.org/Public_Comment_Docs/Documents/2011-03/WS-Agreement-Negotiatio n+v1.0.pdf and is due to be released shortly thereafter, pending final OGF Standards Council review upon completion of this public comment period.
Thanks, Alan
P.S.: As a reminder to the group, the European SIENA roadmap current document corresponding to this NIST one has been released; please see http://sienainitiative.eu where you can create a free account to download this or other relevant documents. In particular you may be interested in the presentations at the recent CloudScape III meeting, which was well- attended by European distributed computing infrastructure projects and government representatives, as well as by Dawn Leaf. I include below a couple of direct links to the roadmap document and to the OCCI presentation at that workshop, in case these might be useful.
SIENA Roadmap document:
OGF OCCI Presentation:
Bonus link: OGF 31 Standards Plenary summary talk (March 23) summarizing status of recently released documents:
Alan Sill, Ph.D Senior Scientist, High Performance Computing Center Adjunct Professor of Physics, TTU Vice President of Standards, Open Grid Forum
==================================================================== : Alan Sill, Texas Tech University Office: Drane 162, MS 4-1167 : : e-mail: Alan.Sill@ttu.edu ph. 806-742-4350 fax 806-742-4358 : ====================================================================
------------------------------------------------------------------------
If you would like to unsubscribe from cc_standards, please send an email to listproc@nist.gov with a message body of:
unsubscribe cc_standards
Please note: you must send the message from the account that you are subscribed.
_______________________________________________ occi-wg mailing list occi-wg@ogf.org http://www.ogf.org/mailman/listinfo/occi-wg
_______________________________________________ occi-wg mailing list occi-wg@ogf.org http://www.ogf.org/mailman/listinfo/occi-wg ------------------------------------------------------------- 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.