North American Network Operators Group

Date Prev | Date Next | Date Index | Thread Index | Author Index | Historical

Re: QoS for ADSL customers

  • From: Joe Shen
  • Date: Tue Nov 29 22:08:55 2005
  • Domainkey-signature: a=rsa-sha1; q=dns; c=nofws; s=s1024; d=yahoo.com.sg; h=Message-ID:Received:Date:From:Subject:To:Cc:In-Reply-To:MIME-Version:Content-Type:Content-Transfer-Encoding; b=YYGWBgHYSXzx3/CMpuT8BpqBmKT2EVb2djMZkkxW3/Y0qCUfs2obJJmE4mdv5gwhZ5Ux6SrKM8od9K7lJAh6iV5kZWiQKrf1afiznqnYXB3QlhYK1PEBHDQk6Vrj0D/Qi6kFOZqG3ttMCFEwHSR0wPifJ+UXdcMbY6WCvIr2UUc= ;

> 
> While some people will cry network neutrality and
> think the Yellow Pages
> must sell only one size listing, some people are
> willing to pay for
> differentiated service.  Trying to classify "bad"
> traffic can be
> done using products like Sandvine.  But it may be
> easier to classify "premium"
> traffic and mark it for special handling, and then
> treating everything
> that isn't marked as premium traffic as best effort
> traffic.


That may be a simple method to differentiate service
between customers. considering e2e qos parameter
requirement by different network applications,
multiple service levels are required to supported in
ISP network (  both intra-ISPnetwork and
inter-ISPnetwork).
 

> 
> But expect great wailing and gnashing of teeth over
> setting or changing
> DSCP/TOS bits or creating different queues for
> different traffic.  Should
> DSCP bits in IP headers be treated like TTL bits
> which are modified by
> the network.  Should ISPs use anti-spoofing
> techniques similar to prevent
> the use of arbitrary IP addresses to control
> DSCP/TOS values in packet
> headers?

To Kim's situation, IP packet header based (or access
interface based) traffic classification is pratical.
If application based traffic classification is
required, tools from sandvine or packeteer may have to
be sitted between ERX1440 and Cisco7609. IMHO, ISP
network should NOT trust any TOS/DSCP set by their
customers; so, classifying and (re)tagging must be
done in PE or BRAS. On the other hand, anti-spoofing
configuration must be enabled in ERX1440 or 7609.
Anyway, I don't trust current router's ability on
content based traffic delivery.

> 
> Most routers already give priority to some types of
> traffic, such as
> routing update packets.

Only with routing protocol packets, it's far from what
we need for service differentiation.

Would Kim share his experience with this work?

Joe




		
__________________________________ 
Meet your soulmate!
Yahoo! Asia presents Meetic - where millions of singles gather
http://asia.yahoo.com/meetic