March 2010 - occi-wg - lists.ogf.org

Re: [occi-wg] OCCI Editor Getting Started Guide (docs/README.txt)
by Andre Merzky 06 Apr '10

06 Apr '10

Hi Sam, all, ready to discuss licensing again? :-) here is an update from OGF28: first, we heard rumours that you may show up here - this would be great, not in the least so that we can discuss the licensing thingie F2F. Usually that is way more productive than endless mailing threads like this one, isn't it :) Anyway, I just wanted to tell you that we have been discussing the issues from the recent email exchanges (and I hope sincerely that these are the issues you are in fact concerned about) within the GFSG. So, there was some discussion to review the current IPR text, and to clarify those passages which seem, in your reading, to hide the fact/intent that using any amount of text from a GFD for documentation and any other purpose is legally perfectly fine: "...derivative works that comment on or otherwise explain it or assist in its implementation may be prepared, copied, published and distributed, in whole or in part, without restriction of any kind..." We will be cross-checking with similar texts of other standards bodies, in particular including IETF, which seem to recently have updated their IPR texts as well, to remove some ambiguities. The OGF IPR was originally modeled after IETF's, and, IIRC, was basically reproducing the OGF IPR word by word. Further we have been discussing the point you raised that the state of OGF documents would be frozen in the case of OGF's disappearance. If this turns out to be a concern for the community, we will consider adding a clause which would release the documents into the public domain in the case of OGF's demise - we certainly don't want to hold anybody back in continuing to work with OGF documents in that case. Both of the above changes however need to be evaluated, and need to be approved by the OGF board. While we don't think this is a problem per se, this will need time to be changed. The OGF IPR is designed as it is to fulfil three goals: (a) support the production and consumption of standards, (b) ensure that OGF documents (i.e. documents released under OGF copyright) went through the OGF document process, and (c) secure OGF against legal litigations. The board will need to make sure that in particular (b) and (c) are not affected by the proposed changes. If this sounds overly bureaucratic to you: well, that is the way we work ;-) Please let us know if changes along those lines would make you sleep better :-D FWIW: for the same reasons as above (a-c), we do actually require that OGF documents are under full OGF copyright, and it does not seem likely that a proposal for dual licensing would find much support, if any. Looking forward to see you in Munich! Best, Andre. -- Nothing is ever easy.

10 27

Question about OCCI Model
by Ralf Teckelmann (Ralf.Teckelmann＠hs-furtwangen.de) 31 Mar '10

31 Mar '10

Hello everybody, I have some little questions about the OCCI Model descripted in occi-core.pdf and occi-infrastructure.pdf. I'm looking at the abstract "levels" of this model at the moment. This is what I got from these papers: We have the resources as concrete entities at the lowest level in the tree. These resource are grouped in categories. A Category belongs to a certain scheme. And on the top of the Model is the "Root" allowing multiply schemes. Or in a top-down way: / --> scheme (1..n) --> category (3..n)--> resource (0..n) What I can't adept is: Is it possible to have sub-categories within a category? Let's say I want to build a little structure below the "Compute" category within the (Standard) "Kinds" scheme. For example I have 2 different forms (avoiding the word kinds) of computing resources I want to structure into 2 sub-categories. Can a resource be part of multiple categories? Let's say I have a couple of resources, which are grouped under the "Compute" category, but I also want them to have a place or appear below another category of another scheme. For example I want to have my own scheme reflecting my virtual infrastructure in any way, so my VMs should appear in both schemes. In one as pure VMs and in the other one for example in relation to other resources (e.g. as part of a cluster). Best regards, Ralf

2 1

No Demo Call today
by Mark Carlson 31 Mar '10

31 Mar '10

Folks, We will skip today and meet next week. -- mark -- To unsubscribe, reply using "remove me" as the subject.

1 0

no concall today - can't open bridge - sorry! n/t
by Thijs Metsch 31 Mar '10

31 Mar '10

-- Thijs Metsch Tel: +49 (0)941 3075-122 (x60122) http://blogs.sun.com/intheclouds http://www.twitter.com/befreax Software Engineer Cloud, Grid and Virtualization Sun Microsystems GmbH Dr.-Leo-Ritter-Str. 7 mailto:thijs.metsch@sun.com D-93049 Regensburg http://www.sun.com

1 0

Contact details change
by Thijs Metsch 31 Mar '10

31 Mar '10

Hi @all, Since I'll be leaving Sun (http://bit.ly/bAymsJ) I wanted to forward my new contact details to you all. My new email address will become tmetsch(a)platform.com by 1st of May 2010. Until then my Sun address will still be valid. But please update your address books. All the best, -Thijs

1 0

Job/Scheduling parameters related to energy efficient scheduling in OCCI
by Mehdi Sheikhalishahi 30 Mar '10

30 Mar '10

Dear All: After "energy efficient scheduling in OCCI" presentation in OGF28, I am sending this message to introduce some Job/Scheduling parameters that can be incorporated into OCCI. According to the complexity of interface and requirements, all or some of these can be incorporated into the OCCI reference model. In Energy Efficient Scheduling, energy consumption for running all jobs is another metric to be optimized; an energy-aware scheduler reduces or minimizes energy consumption to run all jobs in a data center by considering other scheduling metrics. Contemporary modern processors offer a few important features to control energy consumption, e.g. cpuidle and Dynamic Voltage and Frequency Scaling (DVFS, cpufreq). Here comes Job/Scheduling parameters and QoS metrics from user's perspective. hpc-workload characteristic: {io-intensive, cpu-intensive, memory-intensive}, useful for scheduling algorithms like Gang programming model: {serial, MPI, openmp, others} scheduling goal: {energy, performance}, which one should be optimized preemption: {yes,checkpoint,no} allocation type: {best-effort, advance-reservation, immediate, deadline}, type resource allocation for job arrival time: wait time: start time: real execution time: estimated execution time: required by Backfilling algorithms finish time: response time CPU time slowdown algorithm: main scheduling algorithm {aggressive-backfilling, conservative-backfilling, probabilistic-backfilling, lookahead-backfilling, fcfs, sjf, gang} frequency-list: a list of frequencies that a job can run with electricity: how much electricity is used to finish job heat: how much heat is generated during job execution suspension/resumption: suspension of Virtual Machine by scheduler After reviewing parameters related stuff, we will define new namespaces to cover these parameters. Any comments appreciated. Regards, Mehdi

2 1

Occi resources without titles
by Gary Mazz 29 Mar '10

29 Mar '10

Hi, In the current specification, we do not require a resource "title" attribute. However, in the examples, the specification show canonic serialized forms of resources, collections and kinds as a URI using identifiers that I interpret as the occi titles. How should we handle resources created without titles. I have a few suggestions: 1) Allow the provider to assign a resource name. Preferably using a scheme of Resource Identifier (storage, compute, network) plus an number with scheme infinitely incrementing for a specific kind collection. ie Storage01, Storage02, Storage03. If "Storage02" is deleted, the new would be "Storage04". 2) Use the resource identifier until the title attribute is set. 3) Leave the title unset and use to resource kind to identify the resource. 4) Force an occi client to submit a resource title with each create action. This may be impractical if the provider creates resources for you. Number one is my favorite... other suggestion welcome.. cheers, gary

3 3

OpenNebula OCCI at the OGF 28
by Tino Vazquez 24 Mar '10

24 Mar '10

Dear group, I think this may be of your interest, please find in [1] the description and slides for the tutorials given in OGF 28 @ Munich regarding OpenNebula and OGF OCCI. Regards, -Tino [1] http://opennebula.org/documentation:tutorials -- Constantino Vázquez, Grid & Virtualization Technology Engineer/Researcher: http://www.dsa-research.org/tinova DSA Research Group: http://dsa-research.org Globus GridWay Metascheduler: http://www.GridWay.org OpenNebula Virtual Infrastructure Engine: http://www.OpenNebula.org

1 0

CloudCamp
by Bjoern Boettcher 24 Mar '10

24 Mar '10

Hi everybody! I am one of the organizers of the Cloud Camp in Hamburg. I thought it is a good idea to have a presentation about the OCCI at the CloudCamp. Do you have any interest in doing a talk? Regards, Bjoern Boettcher

1 0

Reminder: concall today
by Thijs Metsch 24 Mar '10

24 Mar '10

Hi, Just a short reminder for today's concal at 4pm CET, 8am PDT and 3pm GMT...Please note the slightly changed times for US because of the daylight saving time, which happened in the US but not in Europe yet... I'll give an update on stuff which happened last week during OGF and we'll set some action items to work on the specs and work on the comments we got. This has high priority. Cheers, -Thijs -- Thijs Metsch Tel: +49 (0)941 3075-122 (x60122) http://blogs.sun.com/intheclouds http://www.twitter.com/befreax Software Engineer Cloud, Grid and Virtualization Sun Microsystems GmbH Dr.-Leo-Ritter-Str. 7 mailto:thijs.metsch@sun.com D-93049 Regensburg http://www.sun.com

1 0