Transcript Cisco Presentation Guide

QoS
Quality of Service
Presentation_ID
© 1999, Cisco Systems, Inc.
1
Benefits of QoS
QoS features provide improved and more
predictable network service by offering the
following:
• Dedicated bandwidth
• Improved loss characteristics
• Congestion management and Avoidance
• Traffic Shaping
• Prioritization of traffic
2
What makes up QoS?
Quality of Service is managing:
• Loss (packets that never get there)
• Delay (packets that take too long to get there)
• Jitter (variations in the arrival times of packets)
3
Components of QoS - Loss
• Loss refers to the percentage of packets that fail to reach
their destination.
• Loss can result from errors in the network, corrupted frames
and congested networks. With modern switched and
optically based networks corrupted frames and packet
losses due to network noise, interference and collisions are
becoming rare.
• Many of the packets lost in a healthy network are actually
deliberately dropped by networking devices as a means of
avoiding congestion.
• For many TCP/IP based traffic flows, such as those
associated with file and print services, small numbers of lost
packets are of little concern.
• For UDP traffic associated with real-time applications such
as streaming media and voice, retransmission is not feasible
and losses are less tolerable. As a guide, a highly available
network should suffer less than 1% loss and for voice traffic
the loss should approach 0%.
4
Components of QoS - Delay
• Delay or Latency refers to the time it takes for a
packet to travel from the source to the destination.
• Delay is comprised of fixed and variable delays.
• Fixed delays comprise such events as serialization
and encoding/decoding.
(Eg a bit takes a fixed 100ns to exit a 10Mb Ethernet interface)
• Variable delays are often the result of congestion
and include the time packets spend in network
buffers waiting their turn to access the media.
• Delay is a more significant problem for network
traffic that is bi-directional in nature as the delays
tend to be additive.
5
Components of QoS - Jitter
• Delay variation or Jitter is the difference in the delay
times of consecutive packets.
• Jitter results in degraded audio performance.
Jerky motion, loss of video quality or total loss of
video depending on the encoding scheme used.
• Hardware such as IP Phones use a jitter buffer to
smooth out arrival times. However there are limits
on a buffers ability to do this. In general, traffic that
requires low latency will also require that variation
in latency is also kept to a minimum. This is
because any buffering used to reduce jitter will
directly add to the total delay in the network.
• Design rule - voice networks cannot cope with more
than 30 ms of jitter.
6
Components of QoS - Availability
• No point implementing QoS if your network is down!
• Implement redundancy as well.
ISP 2
ISP 1
Multihomed WAN
or Redundant Core Layer
Si
HSRP
Si
STP
7
Quality of Service Requirements for
Data
Use the proven relative priority model to
divide traffic into no more than four
classes, such as:
• Gold (Mission-Critical)
Transactional, software
• Silver (Guaranteed-Bandwidth)
Streaming video, messaging, intranet
• Bronze (Best-Effort and Default class)
Internet browsing, E-Mail
• Less-than-Best-Effort (Optional; higher-drop preferences)
FTP, backups, applications (Napster, KaZaa)
8
Quality of Service Requirements for
Voice
Voice traffic should be given:
• Loss should be no more than 1%.
• One-way latency should be no more than 150-200 ms.
• Average jitter should be no more than 30 ms.
• 21-106 kbps of guaranteed priority bandwidth is
required per call (depending on the sampling rate,
codec and Layer 2 overhead).
• 150 bps (+ Layer 2 overhead) per phone of guaranteed
bandwidth is required for Voice Control traffic.
9
Quality of Service Requirements for
Video
Requirements vary:
• Video conferencing requirements are similar to voice.
• Streaming media is often buffered for several seconds
so latency requirements can be relaxed.
• Allow for video’s bursty nature…
10
Quality of Service Mechanisms
QoS Service Models:
• Best Effort
The default if no explicit QoS is configured
• Integrated Services Model – IntServ
RSVP – A pre-negotiated QoS path is established end-to-end.
Not well established as the application software must do the
negotiating.
• Differentiated Services Model – DiffServ
Each hop (router) prioritises traffic according to configuration.
Sometimes referred to as a per-hop-behaviour.
• DiffServ is the focus of this course
11
Establishing Differentiated Services
There is a need to “tag” traffic with a QoS level
so that a specific per-hop treatment can be
applied:
• Layer 2 – CoS – Class of Service field.
• 3 bits
0-7 value
• Field is present in ISL and 802.1Q/P encapsulations
• Layer 3 – ToS – Type of Service field.
• 3 bits
0-7 value
Only relevant to IP
• Often referred to as “IP Precedence”
• Layer 3 – DSCP – Differentiated Services Code Point
• Supersedes ToS
• 6 bits (first 3 bits are ToS)
0-63 value
0 is the lowest priority
12
Establishing Differentiated Services
It may seem confusing to have three
options for marking traffic:
• The way to proceed is often determined by the QoS
capabilities of hosts, switches and routers within the
network.
• In many instances it may be necessary to use different
marking techniques at different points within a
network.
• For example, it is common to use the DSCP to mark the
QoS requirements of packets through the routed layers
of the network and mark the frames using the CoS to
allow layer 2 devices such as switches to provide for
the QoS requirements of packet at the data link layer.
13
Modular QoS Command Line Interface
MQC
The Modular QoS Command Line Interface or
MQC is central to Cisco’s model for
implementing IOS based QoS solutions. The
MQC breaks down the tasks associated with
QoS into modules that:
• Identify traffic flows
• Classify traffic flows as belonging to a common class
of QoS.
• Apply QoS policies to that class
• Define the interfaces on which the policy should be
enforced.
14
MQC Classification of traffic
The class-map
Catalyst
2950
Catalyst
3550
Description
access-group
X
X
Access group
ip dscp
X
X
A specific DSCP
value or a list of
values
ip precedence
X
A specific IP
precedence value or
a list of values
any
X
Any packet
class-map
X
A nested class-map
destinationaddress
X
A destination MAC
address
input-interface
X
Select an input
interface to match on
mpls
X
Multiprotocol Label
Switching values
protocol
X
Match on protocol
type
source-address
X
A source MAC
address
vlan
X
VLAN ID
Match on
• The class map is used to associate
one of several attributes with a QoS
treatment that should be given to
that traffic.
• The attributes available vary
depending on the hardware
platform.
Example:
Switch(config)# class-map match-any critical
Switch(config-cmap)# match interface fastethernet 0/1
Any traffic coming in on fa0/1 will be classed as “critical”.
In the next slide we define what to do with “critical” traffic.
15
MQC - Defining the QoS Policy
The policy-map
The policy-map command is used to create a traffic
policy. The purpose of a traffic policy is to configure
the QoS features that should be associated with the
traffic that has been classified in a user-specified
traffic class. A traffic policy contains three elements:
• Policy Name
• Traffic class (specified with the class command)
• QoS policies to be applied to each class
Eg.
Switch(config)# policy-map policy1
Switch(config-pmap)# class critical
Switch(config-pmap-c)# bandwidth 3000
A bandwidth total of 3000 kbps will be given to the traffic classified as “critical” by
the class-map in the previous slide.
16
MQC - Applying the policy to an interface
The service-policy
Just like an Access List, you must apply a
service-policy to a particular interface and
specify it as applying to input or output traffic.
Switch(config)# interface fastethernet 0/1
Switch(config-if)# service-policy output policy1
Switch(config-if)# exit
17
MQC – Using QoS in real networks
When classifying traffic, several common
situations arise:
• Downstream hardware has already set the QoS field at layer 2 or 3.
For example an IP phone can set CoS.
– You can choose to “trust” the device and copy the ToS/CoS.
– You can re-write the CoS to a new value
– Perhaps trust an IP phone but not the CoS/ToS from a PC…
• Downstream device has not set any QoS field.
– You can assign traffic with a default QoS value.
• At the edge of the network (Access layer, and links to other
autonomous systems) it is common to not trust CoS/ToS values.
Rather you would use ACL’s to define the QoS requirements and write
an appropriate CoS/Tos Value
• In the core of your network it is likely that you will trust the CoS/ToS
that you assigned at the edge.
18
Winners and Losers
In order to give priority QoS to one class of
traffic, another (lower) class of traffic must
suffer.
• Policing – Involves either marking down the DSCP value for
packets that are exceeding the bandwidth allocation (non
conforming) or dropping the packet.
• Policing uses a
“Token Bucket” to
determine non
conformance.
• Rate and burst-size
are configurable
• CAR – Committed Access
Rate – Similar to Frame
Relay’s CIR
19
Scheduling
Scheduling is used to give different
queueing priorities according to the
packet or frames DSCP or CoS value.
• First In First Out – FIFO (No QoS treatment)
• Weighted Fair Queuing - WFQ
• Class Based Weighted Fair Queuing - CBWFQ
20
Weighted Fair Queuing (WFQ)
WFQ Services queues with a higher ToS more frequently than those
with a lower ToS
Traffic Destined
for Interface
Transmit Queue
Output Line
Classify
Weighted Fair Scheduling
Configurable Number of
Queues
Flow-Based Classification by:
•Source and destination address
•Protocol
•Session identifier (port/socket)
Interface Buffer
Resources
Weight Determined by:
•Requested QoS (IP Procedure, RSVP)
•Frame Relay FECN, BECN, DE
(For FR Traffic)
•Flow throughput (weighted-fair)
21
Class-Based Weighted Fair Queuing
Class 1
Traffic is
grouped into
user-defined
classes.
BW=64
4
Weight=32
3
2
2
Class 2
BW=128
3
4
Interface
Weight=16
1
2
1
1
4
3
WFQ
Dispatch
Class 3
BW=32
Weight=64
• The weighted fair queuing algorithm is applied
to classes rather than the flows themselves.
© 2002, Cisco Systems, Inc. All rights reserved.
BCRAN v1.1—13-16 22
Low Latency Queuing
Priority Class
V V
1 1
Class 1
2 2
Class 2
Class 3
Class-Default
3
3
3 3
4
4
4 4
7
6 5
Interface
PQ
4
3
2
V V
1 1
WFQ
• LLQ provides for strict priority queuing of voice
traffic (V).
© 2002, Cisco Systems, Inc. All rights reserved.
BCRAN v1.1—13-27 23
Congestion Avoidance
Prioritising traffic in a congested network is fine
but it would be better to avoid the congestion
altogether.
• Congestion leads to dropped packets. By default packets are
dropped indiscriminately once a router’s buffers are full.
This is known as “tail drop”.
• Dropping packets causes TCP to reduce its window-size
thus reducing the data rate and lessening congestion –
good!
• Tail drop causes many TCP sessions to do this
simultaneously – bad!
•This means that bandwidth may not be fully utilised and it
results in a traffic flow that resembles a “saw tooth”.
•Tail drop can result in bursty traffic flows that cause other
problems such as jitter.
24
Congestion Avoidance
Weighted Random Early Detection - WRED
RED drops packets randomly when a routers buffer fills
beyond a certain threshold. This random dropping
prevents the problems associated with tail-drop. No
actual data loss occurs because TCP retransmits.
WRED takes this one step further and facters in QoS
parameters when it “randomly” drops a packet.
WRED drops packets according to the following criteria:
• RSVP flows are given precedence over non-RSVP flows, to ensure that
time-critical packets are transmitted as required.
• The IP precedence or DSCP value of the packets. Packets with higher
precedence are less likely to be dropped. You can control how WRED
determines when and how often to drop packets based on precedence
value if you are not satisfied with the default settings.
• The amount of bandwidth used by the traffic flow. Flows that use the
most bandwidth are more likely to have packets dropped.
• The weight factor you have defined for the interface determines how
frequently packets are dropped.
25
Traffic Shaping
• Designed to smooth traffic flows by limiting
instantaneous bandwidth.
• Cisco IOS QoS software has three types of traffic shaping:
• Generic Traffic Shaping (GTS)
• Class-based (CAR)
• Frame Relay Traffic Shaping (FRTS)
• All three traffic shaping methods are similar in implementation,
though their CLIs differ somewhat and they use different types of
queues to contain and shape traffic that is deferred. In particular
the token bucket is employed by all schemes.
26
QoS over low speed links
• The limited bandwidth of a low speed WAN
link creates particular challenges on QoS.
• Prioritisation of delay sensitive traffic
becomes critical.
• Sometimes waiting for just one packet is
too long!
27
QoS over low speed links
• Examine the table below. It outlines the time taken
for various sized packets to be sent over different
bandwidth WAN links.
• If we have a 64kbps ISDN link and our voice packet
is in the queue behind a 1500 byte FTP packet we
have a problem because the maximum latency for
voice is around 200mS. This doesn’t give us much
to play with if there are several more WAN hops for
our voice packet to traverse.
1 Byte
64
Bytes
128 Bytes
256
Bytes
56 kbps
143 us
9 ms
18 ms
36 ms
72 ms
144 ms
214 ms
64 kbps
125 us
8 ms
16 ms
32 ms
64 ms
126 ms
187 ms
128 kbps
62.5 us
4 ms
8 ms
16 ms
32 ms
64 ms
93 ms
256 kbps
31 us
2 ms
4 ms
8 ms
16 ms
32 ms
46 ms
512 kbps
15.5 us
1 ms
2 ms
4 ms
8 ms
16 ms
32 ms
768 kbps
10 us
640 us
1.28 ms
2.56 ms
5.12 ms
10.24 ms
15 ms
1536 kbps
5 us
320 us
640 us
1.28 ms
2.56 ms
5.12 ms
7.5 ms
512 Bytes 1024 Bytes 1500 Bytes
28
Link Fragmentation and
Interleaving LFI
• The solution is to use LFI to cut-up the larger packets so that the
priority packets can be interleaved with fragments of larger packets.
• You can specify the maximum amount of time in milliseconds that
each fragment can take. This will determine the size of the fragment
• LFI uses PPP to handle the reassembly
29
Compressed Real-Time Transport
Protocol
• RTP is the Internet-standard protocol for the transport of real-time
data, including audio and video.
• Compressed Real-Time Transport Protocol, or CRTP, is used on a
link-by-link basis to compress the IP/UDP/RTP header. In a packet
voice environment when framing speech samples every 20
milliseconds; this scenario generates a payload of 20 bytes. The
total packet size comprises an IP header (20 bytes), a UDP header (8
bytes), and an RTP header (12 bytes) combined with a payload of 20
bytes.
• It is evident that the size of the header is twice the size of the
payload. When generating packets every 20 milliseconds on a slow
link, the header consumes a large portion of the bandwidth. To avoid
the unnecessary consumption of available bandwidth, CRTP is used
on a link-by-link basis.
• This compression scheme reduces the IP/UDP/RTP header to 2
bytes most of the time when no UDP checksums are being sent or 4
bytes when UDP checksums are used.
• Note: Cisco only recommends using cRTP with links lower than 768
Kbps, unless the router is running at a low CPU utilization rate.
Monitor the router's CPU utilization and disable cRTP if it's above
75%.
30
31
Questions
?
32