Re: [clouds] Scope of the Cloud is too big

[Sam Johnston <[e-mail]>]

Gene,

You're definitely not alone in thinking this would be a worthy topic for
IETF and as co-conspirator (along with Cisco's Chris Hoff) and author of the
existing CloudAudit draft spec I'd love to see IETF people starting to work
with the group. We have weekly calls on Monday at 10am PST/1pm EST and
discuss in the cloudaudit Google Group[1]. The goal for v1 is to create a
trivial HTTP-based interface that can be implemented by uploading files
rather than coding. Similarly, OGF's OCCI working group has produced a
number of Internet-Drafts that could be picked up by IETF.

Another area that I think would be very interesting for IETF to take on
(moreso than virtualisation management) is a simple, HTTP-based storage API
like Amazon's S3 - only without the potential patent problems[2]. I believe
that HTTP already takes care of many/most of the issues (e.g.
authentication, encryption, ranged GETs, etc.) and IETF has proven
experience in the area (WebDAV).

Another issue I ran into while writing cush[3] was how to remotely instruct
servers to migrate (live?) resources - for example, moving a virtual
machine, database, etc. from one location to another using a mobile device
on a 3G connection. I believe WebDAV's COPY and MOVE verbs are a good start
(this is what we're using for OCCI) but they could be reviewed and possibly
promoted for more generic application.

Considering that most of these APIs (at least the ones I'm involved in) are
trying to be as close as possible to the "uniform interface" of HTTP,
I believe there's benefit to be had in reviewing the relevant RFCs with
these new applications in mind. For example, while HTTP has a perfectly good
metadata channel (headers) and thus obviates the need for envelope formats
(Atom, SOAP, etc), it lacks the ability to link, annotate and categorise
resources. It also has some internationalisation problems (e.g. ASCII) and
inefficient serialisation (e.g. SPDY). HTTP 1.1 has served us well for many
years but perhaps it's time to start thinking about what HTTP 2.0 might look
like?

Sam

-- 
*Sam Johnston*
*Technical Program Manager*
Site Reliability Engineering
Google Switzerland GmbH

1. http://groups.google.com/group/cloudaudit
2. http://tinyurl.com/s3patent
3. http://code.google.com/p/cush/

On 8 April 2010 15:40, Gene Golovinsky <[e-mail]> wrote:

>  http://www.cloudaudit.org/ is an initiative, not a standard.
>
> The plan for the group is to submit proposal to the IETF:
>
>
> http://searchsecuritychannel.techtarget.com/news/interview/0,289202,sid97_gci1508024,00.html
>
>
>
> Which means I am not alone thinking this is a worthy topic for IETF to take
> on.
>
> When and if the proposal will be submitted there is still going to be a lot
> of work to make it a standard.
>
>
>
> At least two approaches are possible. 1. Do nothing while waiting for Cloud
> Audit proposal. 2. Start working with the group.
>
> Considering how fast Cloud technologies have been evolving and how critical
> Security is for the adoption of it I am for #2.
>
>
>
> --Gene
>
>
>
>
>
>
>
> *From:* Mark Webb [mailto:[e-mail]]
> *Sent:* Thursday, April 08, 2010 8:23 AM
> *To:* [e-mail]
> *Cc:* Linda Dunbar; [e-mail]; Gene Golovinsky
>
> *Subject:* Re: [clouds] Scope of the Cloud is too big
>
>
>
> Look at
>
>
>
> http://www.cloudaudit.org/
>
>
>
> For cloud audit.  It is difficult to specify audit when the services are so
> different today.
>
>
>
> The industry has not settled on a small set of services definitions of what
> "cloud" is. So, how does one audit SaaS, PaaS?  When the services offered in
> that space are so different?  IaaS is the most mature and perhaps the lcd of
> cloud.
>
>
>
> So, who here thinks that "IT functions as a service" or ITaaS is something
> that IETF can specify?  OK that was rhetorical.  My point is, the real
> opportunity is to look for elements that are mature enough to have some
> problem to be solved defined.  Then ensure you are not duplicating what
> other SDO or Forum are already working on.
>
>
>
> Mark Webb
>