Download Weighted fair queuing

Managing Network
Performance
Queuing & Compression
Queuing overview
• A queuing policy helps network managers
meet two challenges:
• Providing an appropriate level of service
for all users
• Controlling expensive WAN costs
Weighted fair queuing
• Weighted fair queuing (WFQ) is a flow-based algorithm
that schedules delay-sensitive traffic to the front of a
queue to reduce response time, and also shares the
remaining bandwidth fairly among high-bandwidth flows.
• By breaking up packet trains, WFQ assures that lowvolume traffic is transferred in a timely fashion.
• Weighted fair queuing gives low-volume traffic, such as
Telnet sessions, priority over high-volume traffic, such as
File Transfer Protocol (FTP) sessions.
• Weighted fair queuing gives concurrent file transfers
balanced use of link capacity.
• Weighted fair queuing automatically adapts to changing
network traffic conditions.
Weighted fair queuing
•
1.
2.
3.
4.
•
There are four types:
Flow-based
Distributed
Class-based
Distributed class-based
Weighted fair queuing is enabled by
default for physical interfaces whose
bandwidth is less than or equal to T1/E1
Conversations or Flows
•
1.
2.
3.
4.
5.
The sorting of traffic into flows is based on
packet header addressing. Common
conversation discriminators are as follows:
Source/destination network address
Source/destination Media Access Control
(MAC) address
Source/destination port or socket numbers
Frame Relay data-link connection identifier
(DLCI) value
Quality of service/type of service (QoS/ToS)
value
Class-based weighted fair
queuing
• CBWFQ extends the standard WFQ functionality
to provide support for user-defined traffic
classes.
• By using CBWFQ, network managers can define
traffic classes based on several match criteria,
including protocols, access control lists (ACLs),
and input interfaces.
• A FIFO queue is reserved for each class, and
traffic belonging to a class is directed to the
queue for that class.
• More than one IP flow, or “conversation", can
belong to a class.
• Once a class has been defined according
to its match criteria, the characteristics can
be assigned to the class.
• To characterize a class, assign the
bandwidth and maximum packet limit.
• The bandwidth assigned to a class is the
guaranteed bandwidth given to the class
during congestion.
• CBWFQ assigns a weight to each
configured class instead of each flow.
• This weight is proportional to the
bandwidth configured for each class.
Weight is equal to the interface bandwidth
divided by the class bandwidth.
• Therefore, a class with a higher bandwidth
value will have a lower weight.
CBWFQ versus flow-based WFQ
• Bandwidth allocation – CBWFQ allows
the administrator to specify the exact
amount of bandwidth to be allocated for a
specific class of traffic.
• The administrator can configure up to 64
classes, and can control distribution
among them.
• Finer granularity and scalability –
• CBWFQ allows the administrator to define
what constitutes a class based on more
criteria.
• CBWFQ allows the use of ACL’s and
protocols or input interface names to
define how traffic will be classified, thereby
providing finer granularity.
• The administrator can configure up to 64
discrete classes in a service policy.
CBWFQ and tail drops
• Traffic variations such as packet bursts or flows
demanding high bandwidth can cause
congestion when packets arrive at an output port
faster than they can be transmitted.
• The router tries to handle short-term
congestions by packet buffering.
• This absorbs periodic bursts of excessive
packets so that they can be transmitted later.
• Although packet buffering has a cost of delay
and jitter, packets are not dropped
• For network traffic causing longer-term
congestion, a router using CBWFQ or any of
several other queuing methods will need to drop
some packets.
• A traditional strategy is tail drop.
• With tail drop, a router simply discards any
packet that arrives at the tail end of a queue that
has completely used up its packet-holding
resources.
• Tail drop is the default queuing response to
congestion.
• Tail drop treats all traffic equally and does not
differentiate between classes of service.
• When using tail drop, the router drops all
traffic that exceeds the queue limit.
• Many TCP sessions then simultaneously
go into a slow start.
• This reduces the TCP window size.
• Consequently, traffic temporarily slows as
much as possible.
• As congestion is reduced, window sizes
begin to increase in response to the
available bandwidth.
• This activity creates a condition called global
synchronization.
• Global synchronization manifests when multiple
TCP hosts reduce their transmission rates in
response to packet dropping, and then increase
their transmission rates after the congestion is
reduced.
• The most important point is that the waves of
transmission known as global synchronization
will result in significant link under-utilization
Weighted Random Early Detect
(WRED)
• WRED monitors the average queue depth
in the router and determines when to
begin packet drops based on the queue
depth.
• When the average queue depth crosses
the user-specified minimum threshold,
• WRED begins to drop both TCP and UDP
packets with a certain probability.
• If the average queue depth ever crosses
the user-specified maximum threshold,
then WRED reverts to tail drop, and all
incoming packets might be dropped.
• The idea behind using WRED is to
maintain the queue depth at a level
somewhere between the minimum and
maximum thresholds, and to implement
different drop policies for different classes
of traffic.
Low Latency Queuing (LLQ)
• The Low Latency Queuing (LLQ) feature
provides strict priority queuing for class-based
weighted fair queuing (CBWFQ), reducing jitter
in voice conversations.
• Configured by the priority command, strict
priority queuing gives delay-sensitive data, such
as voice, preferential treatment over other traffic.
• With this feature, delay-sensitive data is sent
first, before packets in other queues are treated.
• LLQ is also referred to as priority queuing/classbased weighted fair queuing (PQ/CBWFQ)
because it is a combination of the two
techniques.
Queuing comparison
• Flow-based WFQ differs from priority and custom
queuing in several ways.
• On serial interfaces, WFQ is enabled by default, and the
user must enable priority and custom queuing.
• WFQ does not use queue lists to determine the preferred
traffic on a serial interface.
• Instead, the fair queue algorithm dynamically sorts traffic
into messages that are part of a conversation.
• The messages are queued for low-volume
conversations, usually interactive traffic.
• The messages are given priority over high-volume,
bandwidth-intensive conversations, such as file
transfers.
• When multiple file transfers occur, the transfers are given
comparable bandwidth.
• Class-based weighted fair queuing allows
network managers to customize fair queuing
behavior so that user-defined classes of traffic
receive guaranteed bandwidth during times of
congestion.
• More than one flow, or conversation, can belong
to a user-defined class.
• Low latency queuing (LLQ) adds strict priority
queuing to CBWFQ operation.
• LLQ allows a user to specify a priority class,
which will be served before any of the other
classes of traffic. However, the priority queuing
(PQ) with LLQ will not starve the other classes
because the PQ is policed whether or not there
is congestion.
Data Compression
• Data compression works by identifying patterns
in a stream of data, and choosing a more
efficient method of representing the same
information.
• Essentially, an algorithm is applied to the data to
remove as much redundancy as possible.
• The efficiency and effectiveness of a
compression scheme is measured by its
compression ratio, the ratio of the size of
uncompressed data to compressed data.
• A compression ratio of 2:1 (relatively common)
means that the compressed data is half the size
of the original data.
Link compression
• Link compression, which is sometimes
referred to as per-interface compression,
involves compressing both the header and
payload sections of a data stream.
• Unlike header compression, link
compression is protocol independent.
• Uses two types of algorithm:
Predictor
• Predicts the next sequence of characters in the
data stream by using an index to look up a
sequence in a compression dictionary.
• It then examines the next sequence in the data
stream to see if it matches.
• If so, that sequence replaces the looked-up
sequence in a maintained dictionary.
• If not, the algorithm locates the next character
sequence in the index and the process begins
again.
• The index updates itself by hashing a few of the
most recent character sequences from the input
stream.
STAC
• Developed by STAC Electronics, STAC is
a Lempel-Ziv (LZ) compression algorithm.
• It searches the input data stream for
redundant strings and replaces them with
a token, which is shorter than the original
redundant data string.
• If the data flow moves across a point-topoint connection, use link compression.
• In a link compression environment, the
complete packet is compressed and the
switching information in the header is not
available for WAN switching networks.
• Therefore, the best applications for link
compression are point-to-point
environments with a limited hop path.
• Typical examples are leased lines or
ISDN.
Payload compression
• When using payload compression, the
header is left unchanged and packets can
be switched through a WAN packet
network.
• Payload compression is appropriate for
virtual network services such as Frame
Relay, and Asynchronous Transfer Mode
(ATM).
• It uses the STAC compression method
TCP/IP header compression
• TCP/IP header compression subscribes to the
Van Jacobson Algorithm defined in RFC 1144.
• TCP/IP header compression lowers the
overhead generated by the disproportionately
large TCP/IP headers as they are transmitted
across the WAN.
• TCP/IP header compression is protocol-specific
and only compresses the TCP/IP header.
• The Layer 2 header is still intact and a packet
with a compressed TCP/IP header can still travel
across a WAN link.
MPPC
• The Microsoft Point-To-Point Compression
(MPPC) protocol allows Cisco routers to
exchange compressed data with Microsoft
clients.
• MPPC uses an LZ-based compression
mechanism.
• Use MPPC when exchanging data with a
host using MPPC across a WAN link.