Questions about the tman package

Jisu,
comments inline.
-Albrecht

                      "Jisu                            

                      Bhattacharya"            To:     
megacoietf.org                                             

                      <jisu72gmail.co         cc:       

                      m>                       Subject:
[Megaco] Questions about the tman package                   

                      05.04.2007 19:07                      

Hi guys,

I was reading the tman package as defined by ETSI and now
under review as
H.248.53. I have the following questions:

1. Draft H.248.53 makes some suggestions regarding how to
map the tman
parameters into ATM and IP based policing protocols. In that
section, it
uses either pdr + dvt + M OR sdr + mbs + M. So is it
ultimately
recommending either peak-rate policing or sustained-rate
policing but NOT
dual-leaky-bucket policing?

[ABS] Every property-to-policer parameter mapping is
determined by the used
policer algorithm. A general answer on your question is thus
not possible
on your generic question. E.g., if you use GBRA according
Y.1221 for
traffic policing of IP traffic, then there are single- and
dual-LB
algorithms.
So there isn't any ultimative recommendation in H.248.53.

2. What are the valid combinations of pdr, sdr, mbs, dvt and
pol that GC
can specify? For example, can it only specify pdr and mbs?

[ABS] It depends which "traffic characteristic(s)"
you want to policy, e.g.
peak rate, peak rate & sustainable rate, just packet
sizes, etc. There are
many meaningful combinations. The next dependency is again
your selected
policer algorithm.

3. I see tman/pol is now going to default to OFF in
H.248.53. The ETSI doc
does not have a default. Will this cause backward
compatibility issues?

[ABS] Theoretically yes, because ETSI TS 102 333 Annex C was
not explicit
on that subject. Practically, when looking at the H.248
profiles in which
tman/1 is used, I don't think so. When tman/1 was
transferred to ITU-T
H.248.53, we had to formalize the package according the
package template.
The proposal of "OFF" was pending a couple of
meetings without any
objections. But there's still a chance to change that in the
next SG16
meeting. Please come with a correspondent contribution.
I don't have any strong opinion on that. "ON" or
"NONE" might be also
meaningful. Guess "OFF" was considered to be more
conservative.

Thanks,

Jisu Bhattacharya
Cisco
Systems_______________________________________________
Megaco mailing list
Megacoietf.org
https:/
/www1.ietf.org/mailman/listinfo/megaco

_______________________________________________
Megaco mailing list
Megacoietf.org
https:/
/www1.ietf.org/mailman/listinfo/megaco

Questions about the tman package

Thread: Questions about the tman package