Re: [v6ops] new draft: draft-servin-v6ops-monitor-ds-ipv6

Hi all!

My review and proposals:

In the section 1 avoid the use of the word "problem", you can use incident.
Maybe  "Detect and avoid network incidents" fits better.
Same to "Determinate which actions may solve or explain an incident"

In section 2.3 you can include the ipv6 traffic metering based on
forwarding ipv4 and ipv6 traffic through separate paths, and monitoring
them as usual with octets/time counters.

Another technique may be to apply some kind of access-lists or policy-maps
which usually have MIB support.

Regards.

On Mon, Jul 15, 2013 at 8:49 AM, Arturo Servin <aservin@lacnic.net> wrote:

> Diego,
>
>     Thanks for the comments.
>
>     I will try to incorporate some changes before the deadline. Some
> questions in line.
>
> On 7/15/13 6:46 AM, Diego R. Lopez wrote:
> > Hi,
> >
> > I'd say collecting the current state-of-the-art and the operational
> practice in monitoring will be of great help, especially when it comes to
> the tricky inequalities you mention... I really hope to see a comprehensive
> collection of these detail, collected as you say from direct experiences
> across the WG and other fora.
> >
> > Regarding the current contents, I'd like to ask you to include
> considerations for at least:
> >
> > 1) The usage of SDN (OpenFlow in particular) as an alternative
> technique, probably under 2.3
>     I have never thought about it but it seems plausible. Do you have a
> reference, perhaps some previous work about this topic?
>
> >
> > 2) The implications of the usage of local addresses and
> privacy-protection address allocation techniques
>     Noted.
> >
> > And there is always the matter of cross-provider or federated
> monitoring, that I'd say is not that close even in v4 space, and that
> probably would deserve some discussion as well...
>
>     I'll ask some operators that I know have virtual services how they
> do it, but if you have some pointers or references would be very greate
> >
> > Be goode,
> >
> >
>     Thanks for the comments, as I said we will try to incorporate as
> many as possible before the deadline.
>
> Regards,
> as
> _______________________________________________
> v6ops mailing list
> v6ops@ietf.org
> https://www.ietf.org/mailman/listinfo/v6ops
>