Re: [sami] A new draft on state migration use cases is submitted.

Well, I won't try to answer why ISPs and DC providers raise these
requirments. I am sure not all of ISPs and DC providers ask for the same
thing, but we do get the requirements from our customers, at least they can
represent part of the operation views. 

As for your point of "neither Hypervisors nor switches support state live
migration", that is not big deal. Because we make all these discussions here
to figure out whether this is a problem and whether we can resolve it,
instead of whether there is existing solutions for this. 

I think Software Driven Network has no clear relationship to State Migration
after I read the drafts. I may be wrong, could you give some hints based on
your knowledge?

Thanks.

Best Regards
Gu Yingjie

-----邮件原件-----
发件人: sami-bounces@ietf.org [mailto:sami-bounces@ietf.org] 代表 Juergen
Schoenwaelder
发送时间: 2011年10月12日 乐乐14:18
收件人: Yingjie Gu(yingjie)
抄送: 'A tao'; '卓志强(研七 福州)'; sami@ietf.org; '刘茗(研六 福州)'; 'Linda
Dunbar'
主题: Re: [sami] 答复: A new draft on state migration use cases is
submitted.

On Wed, Oct 12, 2011 at 10:57:04AM +0800, Yingjie Gu(yingjie) wrote:

> 2. Some large ISPs we talked with, who are also DC service providers,
> require to move additional network functions to network side, so that they
> can save CPU processing capability to create more VMs, and selling VMs is
> the way they can earn money. 

I assume security functions embedded in network equipment do not come
for free either. So these ISPs seem to believe that security functions
in network devices are easier to manage and cheaper to scale.

> 4. Even for the security functions deployed on Hyperviser, not all of the
> Hypervisor vendors can support state live migration. This is another
reason,
> mentioned by Ming also, why people want to move network functions to
> network.

My understanding is that today neither hypervisors nor 'switches'
support the state live migration you want. Hence, these ISPs seem to
believe that 'switches' are easier to get updated with state migration
functions than hypervisors (even though there is quite some software
out there to manage VM infrastructures).

Since I am not a big ISP, I can't really judge any of this. But those
two arguments at least seem somewhat counter intuitive to me. Anyway,
I like to point your attention to the "Software Driven Networks" BOF
(scroll down on http://trac.tools.ietf.org/bof/trac/wiki/WikiStart).
This might actually be the solution to your problem.

/js

-- 
Juergen Schoenwaelder           Jacobs University Bremen gGmbH
Phone: +49 421 200 3587         Campus Ring 1, 28759 Bremen, Germany
Fax:   +49 421 200 3103         <http://www.jacobs-university.de/>
_______________________________________________
sami mailing list
sami@ietf.org
https://www.ietf.org/mailman/listinfo/sami