GroupStudy.com GroupStudy.com - A virtual community of network engineers
 Home  BookStore  StudyNotes  Links  Archives  StudyRooms  HelpWanted  Discounts  Login
RE: Qos Beginner [7:86262] posted 03/25/2004
[Chronological Index] [Thread Index] [Top] [Date Prev][Date Next] [Thread Prev][Thread Next]


Hello Chuck,

Thanks you for your response and your question.

>And to the person who posted the original question, why does he think he
>needs QoS? Has he done the necessary research, both reading and in terms of
>bandwidth utilization analysis, to make an intelligent determination?

I have not made the detemination yet, Chuck.
To be honest with you, I am trying to learn about QoS more,
to evaluate what's required to implement QoS,
the benefits and downside of QoS and so on,
so that I can ultimately make the decision of implementing QoS or not.

I also have replied the below in my previous post.
===============================
A lot of people raised up the question.
'Do I really need Qos?
I also thought about this question before I started reading.
And I decided to seriously consider implementing Qos,
based on the below reasons.
1. I have VoIP traffic(Not so extensive yet, but will be growing), that I
feel that I should give priority to.
2. My company may start doing some e-learning, which will required
video/audio multicast traffic,
in which case, Qos will come in handy. (based on my shallow knowledge of Qos)
3. I also figures Qos can decrease the damage of certain network
problems/outage.
For instance, I figured it can prevent worms like Blaster and SQL slammer
from completely and spontaneously knocking out my network.
===================================

Granted, my network working smoothly at the moment,
but I can see the traffic in my network steadily growing next year or two,
but I do not see myself getting a whole new set of switches and routers.
So I figured it would be a good idea to look into QoS.

Anyway, based on what I explained above, 
do you think I will benefit a lot from implementing QoS?
Considering the time and NRG required to implement and manage it?

Sean


-----Original Message-----
From: "Chuck Whose Road is Ever Shorter" [mailto:nobody@xxxxxxxxxxxxxx]
Sent: Thursday, March 25, 2004 2:09 PM
To: cisco@xxxxxxxxxxxxxx
Subject: Re: Qos Beginner [7:86262]


Howard, while I defer to your superior intellect, may I ask a question in
line below?


""Howard C. Berkowitz""  wrote in message
news:200403250446.i2P4kCM0025664@xxxxxxxxxxxxxxxxx
> At 4:38 AM +0000 3/25/04, "Chuck Whose Road is Ever Shorter" wrote:
> >actually, I've read various Cisco QoS chapters twice as part of
> >certification test preparation. I appreciate that Cisco indeed lumps
> >policing and rate limiting into the QoS universe, if for no other reason
> >than there is no other place to put them.
> >
> >I was trying to keep this on a conceptual level.
>
> And I think it works conceptually if you distinguish between
> proactive and reactive methods. Most people, I think, will put RED
> into proactive, but there seems to be more of a problem putting
> connection admission control into QoS. Yet connection-oriented CAC is
> the fundamental mechanism of ensuring QoS in the PSTN. Oh, there are
> special compression techniques, but fundamentally things work by
> reserving bandwidth.
>
> Shaping and policing are proactive in a different way, in that they
> prevent the known resources from becoming overloaded and thus
> congested. Now, in a connectionless environment, you may reasonably
> get away with some statistically defined oversubscription, especially
> when the core tends to be overprovisioned.

I look at policing and access rate control not as proactive at all. In terms
of how it is used, sold, conceived of, how is it "proactive" in terms of
buffer management as opposed to something neither proactive or reactive, but
rather a means of limiting bandwidth for economic, sociological, or
philosophical reasons. For example, many instutions of higher learning rate
limit bandwidth to dormatories or affilated institutions despite having
backbones in the multiple gigabit ranges. To them, bandwidth isn't scarce at
all. The reason for rate limiting may be something entirely unrelated.

To get back to my original and primary point, when there is sufficient
bandwidth, QoS is irrelevant and unnecessary. QoS is effectively a more
sophisticated method of queue management, in my simple mind, anyway?

And to the person who posted the original question, why does he think he
needs QoS? Has he done the necessary research, both reading and in terms of
bandwidth utilization analysis, to make an intelligent determination?



>
> >
> >
> >""Jonathan Hays""  wrote in message
> >news:200403242022.i2OKMIge024578@xxxxxxxxxxxxxxxxx
> >>  >Not up to speed on RSVP, so can't comment intelligently. RSVP
> >>  >considers end
> >>  >to end bandwidth availability, correct?
> >>  >
> >>  >RED and WIRED, IRC, kick in as the transmit buffers begin to fill. My
> >>  >understanding of the QoS structure ( on Cisco equipment,
> >>  >anyway ) is that it
> >>  >is a way of managing buffers. Policing / rate limiting is a horse of
a
> >>  >different color, so maybe should be discussed separately from QoS?
> >>  = = =
> >>
> >>  I would agree with you if we were on a generic networking forum. But
> >>  since this is a Cisco forum you might want to peruse the contents of
the
> >>  link below, from which I have extracted the TOC, to understand why
> >>  posters include policing as QoS. Basically anything that even vaguely
> >>  fits into the category of 'QoS' is included by Cisco, and this is true
> >>  for other Cisco publications. The weighty Cisco Press tome "Cisco
DQOS"
> >>  (ISBN 1-58720-058-9; copyright 2004) certainly covers policing.
> >>
> >>
http://www.cisco.com/univercd/cc/td/doc/product/software/ios123/123cgcr/
> >>  qos_vcg.htm
> >>
> >>  Overview
> >>  Part 1: Classification
> >>  Part 2: Congestion Management
> >>  Part 3: Congestion Avoidance
> >>  Part 4: Policing and Shaping
> >>  Part 5: Signalling
> >>  Part 6: Link Efficiency Mechanisms
> >>  Part 7: Quality of Service Solutions
> >>  Part 8: Modular Quality of Service Command-Line Interface
> >>  Part 9: Quality of Service Device Manager
> >>  Part 10: AutoQoS
> >>  **Please support GroupStudy by purchasing from the GroupStudy Store:
> >>  http://shop.groupstudy.com
> >>  FAQ, list archives, and subscription info:
> >http://www.groupstudy.com/list/cisco.html
> **Please support GroupStudy by purchasing from the GroupStudy Store:
> http://shop.groupstudy.com
> FAQ, list archives, and subscription info:
http://www.groupstudy.com/list/cisco.html
**Please support GroupStudy by purchasing from the GroupStudy Store:
http://shop.groupstudy.com
FAQ, list archives, and subscription info:
http://www.groupstudy.com/list/cisco.html



------------------------------------------------------------------------------
The information in this e-mail is confidential and may be legally
privileged. It is solely for the intended addressee. Access to this e-mail
by anyone else is unauthorized. If you are not the intended recipient, any
disclosure, copying, distribution or any action taken or omitted to be taken
in reliance on it is prohibited and may be unlawful. KPMG does not
guarantee, warrant or represent that this email or any attachments will be
error free, virus free, nor that the transmission will not be delayed,
intercepted or interfered.




Message Posted at:
http://www.groupstudy.com/form/read.php?f=7&i=86337&t=86262
--------------------------------------------------
**Please support GroupStudy by purchasing from the GroupStudy Store:
http://shop.groupstudy.com
FAQ, list archives, and subscription info: http://www.groupstudy.com/list/cisco.html