June 2009 - occi-wg - lists.ogf.org

Opinion Poll: IaaS or PaaS ?
by Gary Mazz 03 Jul '09

03 Jul '09

Sam Johnston's timely IaaS feature matrix brings up some interesting issues, one in particular, what are the specific features that can be included in an IaaS. Many of the IaaS provider are also providing one or more operating systems while other are providing closer to bare metal. Is the OS part of the Infrastructure or part of the Platform ? The question is: " Do operating systems belong in IaaS or PaaS. " Please lets us know you opinion -cheers gary mazz

9 30

Reminder: weekly OCCI concall - tomorrow Wednesday July 1st at 5pm CDT...
by Thijs Metsch 01 Jul '09

01 Jul '09

Folks, Again it is time for our weekly concall tomorrow @5pm CDT. Please lookup your timezone and the numbers of the conference bridge at: http://forge.ogf.org/sf/wiki/do/viewPage/projects.occi-wg/wiki/Teleconferen… Currently the agenda is set to: 1. Current status of all work packages * Requirements * Use Cases * Members * API review * Wiki structure 2. Requirements and Use cases 3. API status 4. draft API If you have any more topics or ideas - feel free to edit the wiki page: http://forge.ogf.org/sf/wiki/do/viewPage/projects.occi-wg/wiki/Concall20090… This is a very important call regarding the requirements use cases and our current status - so please join! All the best, -Thijs -- Thijs Metsch Tel: +49 (0)941 3075-122 (x60122) http://blogs.sun.com/intheclouds Software Engineer Grid Computing Sun Microsystems GmbH Dr.-Leo-Ritter-Str. 7 mailto:thijs.metsch@sun.com D-93049 Regensburg http://www.sun.com

1 1

New Discussion Group on High Performance and Grid Computing in the Cloud
by Ignacio Martin Llorente 29 Jun '09

29 Jun '09

Dear OCCIers, I know that many of you are highly interested in the application on Cloud to enhance HPC and Grid computing infrastructures. We have created a google group "HPCcloud: High Performance and Grid Computing in the Cloud" (http://groups.google.com/group/HPCcloud) The purpose of this group is not to define standards but to present experiences and scenarios by individuals, organizations and projects to illustrate how Cloud computing can enhance the different types of distributed and high performance computing infrastructures in science and engineering. I hope that as result of the discussion we will contribute new use cases to OCCI-WG. The group covers the following aspects about innovative potential, benefits and challenges of new Cloud technologies and services in High Performance Computing (HPC) and Grid Computing research and business: - Cultural, security, political and legal barriers to implementing Cloud provisioning models in HPC and Grid environments - Architectures for integration of Cloud technologies and services with HPC and Grid infrastructures - Standardization of interactions between HPC and Grid platforms and Cloud infrastructures - Limitations of existing Cloud services and technologies for the capability and capacity computing demands of the HPC and Grid communities in the execution of both tightly-coupled HPC and loosely- coupled HTC applications - HPC Clouds offering platforms with HPC devices and configurations, and Scientific Clouds offering specific services for the scientific and technical computing community - Impact of virtualization on the performance of memory, CPU and I/O intensive, and latency sensitive applications, and virtualization support for specialized communication transports - Service and infrastructure scalability and elasticity management for the efficient execution of virtualized HPC and Grid platforms - Challenges of porting HPC applications to the Cloud and new computing paradigms for HPC on Cloud Ignacio -- Ignacio M. Llorente, Full Professor (Catedratico): http://dsa-research.org/doku.php?id=people:llorente DSA Research Group: web http://dsa-research.org and blog http://blog.dsa-research.org Globus GridWay Metascheduler: http://www.GridWay.org OpenNebula Virtual Infrastructure Engine: http://www.OpenNebula.org

1 0

Templates (was Re: OCCI Dashboard)
by Sam Johnston 28 Jun '09

28 Jun '09

On Thu, Jun 25, 2009 at 4:44 PM, <shlomo.swidler(a)gmail.com> wrote: > On Thu, Jun 25, 2009 at 5:34 PM, Sam Johnston<samj(a)samj.net> wrote: > > On Thu, Jun 25, 2009 at 3:45 PM, <shlomo.swidler(a)gmail.com> wrote: > >> > >> 4. How will I be able to discover VM images that are appropriate for > >> my "compute"s? For example, Amazon has public AMIs that I can launch > >> on my instances, and I can filter the ElasticFox view to only show me > >> 64-bit AMIs or otherwise filter them by AMI name. Is this use case > >> covered by OVF, or is it more appropriate at the OCCI level? > > > > This one I have a better answer for, in that there should be no "start" > verb > > for a template (e.g. AMI), rather only a "clone" verb which returns a > > Location: to a new compute resource (pre-linked to any storage and > network > > resources as necessary). We could also bundle templates into a > pre-defined > > category, such that machines that are just not startable (e.g. due to > > permissions, faults, etc.) don't appear as templates. > > Sounds strange to me. > > Sheep aSheep = new Sheep(); > Object dolly = aSheep.clone(); > > I would expect "dolly" to be of type Sheep, the same type as the > object that was cloned. > The original object was a compute resource and the result of "cloning" it is a compute resource too - I don't see a problem here (though I agree with your argument). > Perhaps the verb "instantiate" is more appropriate to describe the act > of "using a template". But it's harder to spell. Perhaps, and you're right it is harder to spell. Whatever we choose we just need to make sure we define it properly. For performance reasons it might be useful to be able to "start" a template anyway (one call instead of two)... in which case a pre-defined "template" category makes sense. Sam

3 11

OCCI Dashboard
by shlomo.swidler＠gmail.com 25 Jun '09

25 Jun '09

As an end-user, I'm hoping to eventually see an OCCI-aware ElasticFox (or similar cloud infrastructure provider dashboard) that can be used to view and manage OCCI deployments. (This may just map to a collection of use cases.) I'm comparing the current features of ElasticFox to the features exposed by the proposed API [1] and I note the following about the API: 1. There is no Noun or Attribute for "Security Groups". The API should allow me to define a network security group (consisting of a collection of protocol+port+CIDR specifications), and allow each network interface to belong to multiple security groups. There was some discussion back in May [2] about networking attributes, but it did not get very far. 2. Amazon has distinct nouns for "Volumes" and "Snapshots". Amazon's snapshots are not mountable, you can only "copy"/"clone" (not sure which OCCI verb is the correct one here) an Amazon EBS snapshot into a new EBS drive. The OCCI API seems to imply that the result of the verb "image" on a drive yields another drive. So there may be a need to introduce either a new noun "drive snapshot" that is not linkable to any compute noun, or to otherwise differentiate between drives that are mountable and drives that are not mountable. 3. Similar to #2, but more generally: how can I, the OCCI consumer, know a-priori (without actually doing it) whether the drive that results from "image"ing an existing drive will be linkable (mountable) to my compute resources? Because some providers may produce a full-fledged drive as a result, and some may produce an unmountable "snapshot". Or, put even more generally, is there some meta-information I can probe at the OCCI API level to discover the noun/verb/attribute combinations that are supported by a particular implementation? 4. How will I be able to discover VM images that are appropriate for my "compute"s? For example, Amazon has public AMIs that I can launch on my instances, and I can filter the ElasticFox view to only show me 64-bit AMIs or otherwise filter them by AMI name. Is this use case covered by OVF, or is it more appropriate at the OCCI level? Thanks. .. Shlomo [1] http://forge.ogf.org/sf/wiki/do/viewPage/projects.occi-wg/wiki/APIDesign [2] http://www.ogf.org/pipermail/occi-wg/2009-May/000633.html

2 5

Storage Verbs (was Re: OCCI Dashboard)
by Sam Johnston 25 Jun '09

25 Jun '09

On Thu, Jun 25, 2009 at 3:45 PM, <shlomo.swidler(a)gmail.com> wrote: > 2. Amazon has distinct nouns for "Volumes" and "Snapshots". Amazon's > snapshots are not mountable, you can only "copy"/"clone" (not sure > which OCCI verb is the correct one here) an Amazon EBS snapshot into a > new EBS drive. The OCCI API seems to imply that the result of the verb > "image" on a drive yields another drive. So there may be a need to > introduce either a new noun "drive snapshot" that is not linkable to > any compute noun, or to otherwise differentiate between drives that > are mountable and drives that are not mountable. > We plan to expose a basic set of well-defined nouns which implementors are able to use in an interoperable fashion. This will include basic functionality like clone (possibly with parameters like "full" vs "cow") but things like "format", "check", "backup", etc. may or may not be specified. Note that a "mount" is currently done by creating an association between a compute and a storage resource (specifying the local identifier - eg "sda"). I can't think of a sensible way to advertise "mountability" (short of throwing errors when it fails), and having an "attach" verb on storage devices which accepts a compute device as a parameter feels fugly. Sam

2 2

Firewalling Syntax (was Re: OCCI Dashboard)
by Sam Johnston 25 Jun '09

25 Jun '09

On Thu, Jun 25, 2009 at 3:45 PM, <shlomo.swidler(a)gmail.com> wrote: > 1. There is no Noun or Attribute for "Security Groups". The API should > allow me to define a network security group (consisting of a > collection of protocol+port+CIDR specifications), and allow each > network interface to belong to multiple security groups. There was > some discussion back in May [2] about networking attributes, but it > did not get very far. > My (undocumented) thoughts on this thus far are that we should have an extension that allows users to specify firewall rules on network associations in a simple format like ufw<http://www.ubuntugeek.com/ufw-uncomplicated-firewall-for-ubuntu-hardy.html>or pf <http://www.openbsd.org/faq/pf/filter.html#syntax>: pass in proto tcp from any port www > For "groups" you would associate multiple resources to the same network and then associate that network with another, specifying rules on the association. For more advanced functionality like Netscaler VPX, ZXTM, etc. there would be a dedicated compute or network resource (as appropriate). Feedback welcome, Sam

1 0

Oracle VM Manager Web Services API
by Sam Johnston 25 Jun '09

25 Jun '09

Morning all, Keeping with the one API per provider theme, Oracle have just released a[nother] SOAPy RPC-style API for infrastructure management: Announcing Oracle VM 2.1.5<http://blogs.oracle.com/virtualization/2009/06/announcing_oracle_vm_215.html> Web Services API Docs<http://download.oracle.com/docs/cd/E11081_01/doc/doc.21/e10901/wsapi.htm> It's anybody's guess what those guys are up to, having just taken VirtualIron out the back to shoot it while at the same time acquiring Sun with all its cloud API goodness... oh well at least they're not just selling boxes running VMware :P Sam

1 0

Re: [occi-wg] Ambiguity regarding Collections
by shlomo.swidler＠gmail.com 24 Jun '09

24 Jun '09

Sorry, forgot to CC the list. ---------- Forwarded message ---------- From: <shlomo.swidler(a)gmail.com> Date: Wed, Jun 24, 2009 at 11:43 PM Subject: Re: [occi-wg] Ambiguity regarding Collections To: Sam Johnston <samj(a)samj.net> > Great - we discussed how the specification would be documented on today's > call and there was some concern that it wouldn't be long/formal enough if we > continue on as I've started but I'd prefer to produce something that is as > concise as possible (ideally no more than a few pages). This is yet another reason why it's important to have multiple, independent implementations: spec ambiguity can be discovered by comparing the differences between implementations. Have you considered having an official "OCCI Compatibility Test", which could be an impartial arbiter of whether or not a given implementation is spec compliant? > - If you retrieve a representation of a single resource you will get just > the representation itself (with the metadata out of harm's way in the > headers) > - If you instead request something that *could* return multiple resources > (e.g. a category, search, etc.) then you will receive a feed with zero or > more entries. This seems OK to me, as long as the disposition of each method (whether it only ever returns a single resource, such as GET /compute/uuid1, or whether it can return multiple resources, such as a search) is clearly indicated in the spec. .. Shlomo

2 5

Ambiguity regarding Collections
by shlomo.swidler＠gmail.com 24 Jun '09

24 Jun '09

[I've been a lurker here since the very beginning of this list. I'm an end-user, interested in using OCCI to allow me to minimize the vendor lock-in of my cloud deployments. I have experience in writing and implementing specs of the JSR kind.] There seems to me to be some potential ambiguity in the "spec" regarding collections. I'm referring to the HTML attached [1] to the most recent Conference Call agenda (which I did not attend, so forgive me if this was discussed). The document says: Operations that return multiple resources (e.g. categories, searches) are rendered as an Atom feed with an Atom entry per resource. Does this mean that a conforming implementation can, if there is only a single resource to return, choose to return it as a collection of length 1 OR as a "flat" description without Atom wrapping? Or, does it mean that "operations that, by their nature, can potentially return multiple resources (such as search) will return these results enumerated via an Atom feed REGARDLESS of how many results there are"? .. Shlomo [1] http://www.ogf.org/pipermail/occi-wg/attachments/20090624/01218d83/attachme…

2 1