Transcript Quality of Service (QoS)-Based Management of Preempted Traffic in

Quality of Service (QoS)-Based
Management of Preempted
Traffic in MPLS Networks
Eng. Ayman Maliha
Electrical & Computer Engineering Department
The Islamic University of Gaza
Contents
• Introduction
• MPLS & Traffic Engineering
• Thesis Statement
• Preemption in MPLS (MNS-2)
• Proposed Preemption Technique
• Simulation Results
• Conclusion and Future Work
OSI Reference Model
Source node
Destination node
Application
Application
Presentation
Presentation
Session
Session
Intermediate node
transport
Network
Packets
transport
Network
Network
Data link
Data link
Physical
Physical
Frames
Data link
Physical
Bits
Signals
OSI Reference Model
LAYER
7. Application
6. Presentation
5. Session
ROLE
* Functionality:
+ Implement the desired procedure.
+ Provide the user interface
* Provide enhanced services (Control structure
for communication between applications.)
4. Transport
* end-to-end error recovery and flow control
3. Network
* Provide host to host link
2. Data Link
* Provides for reliable transfer of information
across the physical layer.
1. Physical
* Provide physical connection to the net
OSI Reference Model
Source node
Destination node
Application
Application
Presentation
Presentation
Session
Session
transport
transport
Network
Network
Network
Data link
Data link
Data link
Physical
Physical
Physical
Signals
Intermediate node
OSI Reference Model
Network A Node
7
6
5
Application
AL-Hdr
Presentation
PL-Hdr
Session
4
Transport
3
Network
2
Data Link
1
Physical
SL-Hdr
DLL-Hdr
PL-Hdr
Presentation Layer Msg
Session Layer Msg
TL-Hdr
NL-Hdr
Application Layer Msg
Transport Layer Msg
Network Layer Msg
Data Link Layer Msg
Physical Layer Msg
OSI Reference Model
Source node
Destination node
Application
Application
Presentation
Presentation
Session
Session
transport
transport
Network
Network
Network
Data link
Data link
Data link
Physical
Physical
Physical
Signals
Intermediate node
TCP/IP Reference Model
OSI
TCP/IP
7. Application
Application
6. Presentation
5. Session
Not Present
4. Transport
Transport
Not Present
3. Internet
3.
Network
2. Data Link
2. Data Link
1. Physical
1. Physical
Internet Today
• Internet today
- Provides “best effort” data delivery
- Complexity stays in the end-hosts
- Network core remains simple
- As demands exceed capacity, service degrades
gracefully (increased jitter etc.)
Exceeding the delays and jitter boundaries causes
problems to real-time applications.
Quality of Service (QoS)
• Definition
A set of service requirements to be met by the
network while transporting a flow.
• Goal
Provide some level of predictability and control
beyond the current IP “best-effort” service.
QoS Metrics
- Bandwidth
- Delay (or latency)
- Jitter
- Loss rate
Vary according to Service Level Agreement
(SLA)
QoS Protocol Classification
•QoS
can be achieved by :
–- Resource reservation (integrated services)
–-
Prioritization (differentiated services)
QoS
–-
can be applied :
Per flow (individual, uni-directional streams)
–- Per aggregate (two or more flows having
something in common)
QoS Protocol
IETF
-
Integrated Service (IntServ)
-
Differentiated Services (DiffServ)
-
Multi Protocol Labeling Switching (MPLS)
Integrated Service (IntServ)
• Philosophy Behind
Routers have to be able to reserve resources
to provide special QoS for specific user
packet streams.
• Four components of IntServ Model
• The signaling protocol (e.g. RSVP)
• The admission control routine
• The classifier
• The packet scheduler
IntServ Components
IntServ Components
•The signaling protocol
 Sender sends a PATH Message to the
receiver specifying the characteristics of the
traffic
The receiver responds with a RESV Message
to request resources for the flow
Every intermediate router along the path can
reject or accept the request of the RESV
Message
IntServ Components
• Admission control
Decide whether a request for resources can be
granted
• Classifier
When a router receives a packet, the classifier will
perform classification and put the packet in a
specific queue based on the classification result
• Packet scheduler
Schedule the packet accordingly to meet its QoS
requirements
IntServ Problems
•Problems
– Not scalable
• Huge storage and processing overhead on
the routers
• The amount of state information increases
proportionally with the number of flows
– Requirement on routers is high
• All routers must
admission control,
packet scheduling
implement RSVP,
classification, and
DiffServ
-
Description
 Applied
on flow aggregates
 Services requirements are classified
 Classification is performed at network
ingress points
 A predefined per-hop behavior (PHB) is
applied to every service class
 Traffic is smoothed according to PHB applied
DiffServ functional elements
• edge functions:
– packet classification
– packet marking
– traffic conditioning
• core functions:
– forwarding based on per-hop behavior (PHB)
associated with packet’s class
DiffServ - Traffic Classes
DiffServ functional elements
• packet classification
Classifier selects packets based on values in packet
header fields and steers packet to appropriate
marking function
• Meter
Calculates the traffic level, which is compared
against the customer’s contract/Service Level
Agreement (SLA) profile.
DiffServ functional elements
• Marker
The packets are marked by setting the DS value to
a correct codepoint as needed
• Shaper
The shaper Polices traffic by delaying packets as
necessary so the that the packet does not exceed
the traffic rate specified in the profile for that class
DiffServ functional elements
• Dropper
Drops the packets when the rate of packets of a
given class exceeds that specified in the profile for
that class
• Per-hop behavior (PHB) defines differences
in performance among classes.
DiffServ - Traffic Classes
Two traffic classes are available :
–Expeditied
•
•
•
Minimizes delay and jitter
Provides the highest QoS
Traffic that exceeds the traffic profile
discarded
–Assured
•
•
Forwarding (EF)
is
Forwarding (AF)
4 classes, 3 drop-precedences within each
class
Traffic that exceeds the traffic profile is not
delivered with such high probability
DiffServ - Advantages
• Advantage
– Scalable
• Edge routers maintain per aggregate state
• Core routers maintain state only for a few traffic
classes
– Easy implementation
• Incremental deployment is possible for Assured
Forwarding
DiffServ - Disadvantages
• Disadvantage
– Provide weaker service than InteServ
– per hop behavior cannot guarantee end-to-end QoS.
Multiprotocol Label Switching (MPLS)
3- Multiprotocol Label Switching (MPLS)
 MPLS is a technology that integrates label-swapping
paradigm with network-layer routing within Label
Switching Routers (LSRs).
 A short fixed-length “label” results in high-speed
switching.
 It is proposed to be a combination of the better
properties of ATM and IP.
MPLS
IP Router
Control:
MPLS
Control:
IP Router
Software
IP Router
Software
Forwarding:
Forwarding:
Longest-match
Lookup
Label Swapping
ATM Switch
Control:
ATM Forum
Software
Forwarding:
Label Swapping
Contents
• Introduction
• MPLS & Traffic Engineering
• Thesis Statement
• Preemption in MPLS (MNS-2)
• Proposed Preemption Technique
• Simulation Results
• Conclusion and Future Work
IP Traditional Routing
• Choosing the next hop
 Open Shortest Path First (OSPF) to populate the
routing table
 Route look up based on the IP address
 Find the next router to which the packet has to be
sent
 Replace the layer 2 address
• Each router performs these steps
IP Routing Table
Build IP routing
table
Dest
47.1
47.2
47.3
Dest
47.1
47.2
47.3
Out
1
2
3
Out
1
2
3
1 47.1
3
1
Dest
47.1
47.2
47.3
Out
1
2
3
2
3
2
1
47.2
47.3 3
2
IP Traditional Routing
Traditional IP
forwarding
Dest
47.1
47.2
47.3
Dest
47.1
47.2
47.3
Out
1
2
3
1 47.1
1
Dest
47.1
47.2
47.3
Out
1
2
3
IP 47.1.1.1
2
IP 47.1.1.1
3
Out
1
2
3
2
IP 47.1.1.1
1
47.2
47.3 3
2
IP 47.1.1.1
Disadvantages
• Header analysis performed at each hop
• Increased demand on routers
• Utilizes the best available path
• Some congested links and some underutilized links!
Degradation of throughput
Long delays
More losses
• No QoS
No service differentiation
Not possible with connectionless protocols
MPLS & Traffic Engineering
• MPLS Components
1- Label Switching Based Router (LSR & LER)
A high-speed router device that participate in the
establishment of LSP.
2- Label Switching Path (LSP)
A sequence of LSRs that is to be followed by a packet.
3- Labeled Packets
A packet into which a label has been encoded.
MPLS & Traffic Engineering - (LSP)
Intf Label Dest Intf Label
In In
Out Out
3
0.50 47.1 1
0.40
Intf
In
3
Label Dest Intf
In
Out
0.40 47.1 1
1
Request: 47.1
47.3 3
3
2
3
1
47.1
1
2
Mapping: 0.40
47.2
2
MPLS & Traffic Engineering - (LSP)
Intf Label Dest Intf Label
In In
Out Out
3
0.50 47.1 1
0.40
MPLS Switching
Intf Dest Intf Label
In
Out Out
3
47.1 1
0.50
Intf
In
3
Label Dest Intf
In
Out
0.40 47.1 1
IP 47.1.1.1
1 47.1
3
1
3
2
1
47.3 3
47.2
2
IP 47.1.1.1
2
MPLS & Traffic Engineering - (ER-LSP)
Intf Label Dest Intf Label
In In
Out Out
3
0.50 47.1 1
0.40
Intf
In
3
3
Dest
47.1.1
47.1
Intf
Out
2
1
Label
Out
1.33
0.50
Intf
In
3
Label Dest Intf
In
Out
0.40 47.1 1
IP 47.1.1.1
1 47.1
3
3
2
1
47.3 3
2
47.2
2
IP 47.1.1.1
1
MPLS & Traffic Engineering - Labels
•
•
•
•
A short, fixed length identifier (32 bits)
Sent with each packet
Local between two routers
Can have different labels if entering from different
routers
20
3
1
8
Label Value
Exp.
S
TTL
20-bits : Label value used for lookup
3-bits : Reserved
8-bits : Time-To-Live
1-bit : Bottom of Label Stack
MPLS & Traffic Engineering - Labels
•ATM:
•Frame Relay:
•Ethernet:
•PPP:
Label
Label
Label
Label
VPI/VCI(w/shim)
DLCI(w/shim)
Shim
Shim
MPLS & Traffic Engineering
PPP Header
Layer 3 Header
Label
PPP Header
LAN MAC Header
Layer 3 Header
Label
MAC Header
ATM Cell Header
DATA
HEC
CLP
PTI VCI VPI
Label
GFC
MPLS & Traffic Engineering
MPLS & Traffic Engineering
MPLS & Traffic Engineering
MPLS & Traffic Engineering
TE Definition
An iterative process of network planning and network
optimization
TE Objectives
- High service quality
- Efficiency
- Survivability
- Cost
The established path must fulfill some requirements to deliver the
required QoS and it should satisfy the network capacity and policy.
MPLS & Traffic Engineering.
• TE Attributes of Traffic Trunks
- Traffic parameter attribute i.e. peak rates, burst
size, etc..
- Policing attribute
- Path selection and management attribute
- Priority attribute
- Preemption attribute
- Resource attribute
Preemption
Definition
 Preemption is the premature suspension or
termination of an activity in order to permit some other
activity to proceed.
 It is an action that is taken by a system element when
the demand for the resources exceeds the available supply.
Preemption attribute
 Preemption attribute determines whether a traffic
trunk can preempt another traffic trunk from a given
path.
Adaptive Real-time Traffic
 Adaptive or Controllable real-time applications
can adjust their data rates to the available
bandwidth e.g. videoconferencing (CIF, MPEG-I).
 Such applications could be treated at lower
QoS level that depends on the available
bandwidth.
Contents
• Introduction
• MPLS & Traffic Engineering
• Thesis Statement
• Preemption in MPLS (MNS-2)
• Proposed Preemption Technique
• Simulation Results
• Conclusion and Future Work
Thesis Statement
 Bandwidth allocation is an important issue in
network management dealing with guaranteed
bandwidth policy.
 The ability of an application to maintain its
bandwidth depends on its precedence attribute
within the network.
 Preemption allows guaranteed bandwidth for
high priority traffic.
 Harsh solution for the preempted traffic,
which loses its resources.
Thesis Statement
 Real-time traffic
• Advantageous for the preemptor traffic.
• Disastrous for the preempted traffic
 Adaptive real-time traffic needs to be treated
differently when preempted i.e. serving it at lower bit
rate if the reservable bandwidth meets the new rate.
 Upon preemption, network needs to consider:
• Reservable bandwidth
• traffic type
• priority
Contents
• Introduction
• MPLS & Traffic Engineering
• Thesis Statement
• Preemption in MPLS (MNS-2)
• Proposed Preemption Technique
• Simulation Results
• Conclusion and Future Work
MNS Simulator
 A simulation tool for MPLS network.
 It is implemented as an extension of NS-2
simulator, which is an object-oriented Tcl script
interpreter.
 MNS-2 commands must be written in a Tcl
script file, which defines the simulation
scenario.
Preemption in MPLS
CR-LSP1 300 (kbps)
link
BW
(1Mbps)
Total
Maximum Real-time bw fraction
800400
(kbps)
CR-LSP2
(kbps)
Reservable bandwidth
Best-effort and signaling traffic (200 kbps)
Bandwidth allocation in MNS-2
Time
Preemption in MPLS
CR-LSP2,
CR-LSP1, bw 400
768 kbps
setupPrio 35 , holdprio 2.
4.
link
BW
(1Mbps)
Total
CR-LSP2 (400 kbps)
CR-LSP1 768
CR-LSP1
768 (kbps)
Real-time
bandwidth
fraction (800 kbps)
(kbps)
Reservable bandwidth (32 kbps)
Best-effort and signaling traffic (200 kbps)
Bandwidth allocation for AD-RT in
MNS-2
Time
Preemption in MPLS
 Preempted traffic is served at best-effort level,
and it becomes under the mercy of network load.
 Real-time bandwidth fraction is not well
utilized.
 Preempted real-time traffic sharing other besteffort traffic resources, i.e. no dedicated
resources remain for the preempted traffic.
Contents
• Introduction
• MPLS & Traffic Engineering
• Thesis Statement
• Preemption in MPLS (MNS-2)
• Proposed Preemption Technique
• Simulation Results
• Conclusion and Future Work
Proposed Preemption Technique
Preemption time
reservable bw (16 kbps)
CR-LSP2,
CR-LSP1, bw 400
768 kbps
setupPrio 35 , holdprio 2.
4.
CR-LSP2 (400 kbps)
CR-LSP1
Total link
(786
kbps)
CR-LSP1
768 (kbps)
Real-time
bandwidth
fraction (800 kbps)
BW
(1Mbps)
CR-LSP1 (384 kbps)
Reservable bandwidth (32 kbps)
Best-effort and signaling traffic (200 kbps)
t
Bandwidth allocation for AD-RT in the
proposed preemption mechanism.
Time
Proposed Preemption Technique
Preemption time
CR-LSP1, bw 400
CR-LSP2,
500 kbps
No reservable bw
setupPrio 35 , holdprio 2.
4.
Reservable
Reservable bandwidth (300 kbps)
CR-LSP2 (400 kbps)
bandwidth
link
BW
(1Mbps)
Total
Real-time bandwidth fraction (800 kbps)
CR-LSP1 500 (kbps)
CR-LSP1 (400 kbps)
Best-effort and signaling traffic (200 kbps)
t
Bandwidth allocation for AC-RT in the
proposed preemption mechanism.
Time
Contents
• Introduction
• MPLS & Traffic Engineering
• Thesis Statement
• Preemption in MPLS (MNS-2)
• Proposed Preemption Technique
• Simulation Results
• Conclusion and Future Work
Contents
Simulation results
Total link = 4 Mbit
Bandwidth allocation for two preempted traffics.
RT- fraction = 3520 kbit
RT3
AD-RT1
AD-RT2
)
.
.
.
.
.
.
.
.
(Mbit
Bandwidth
Reservable = 1320 kbit
Time (sec)
Simulation results
Total link = 4 Mbit
Bandwidth allocation for one real-time traffic when
two traffics are preempted
RT- fraction = 3520 kbit
Reservable = 520 kbit
.
RT3
AD-RT1
)
.
AD-RT2
(Mbit
Bandwidth
.
.
Time (sec)
Simulation results
Throughput
• Throughput
AD-RT1
RT1
BET
Background data rate (kbps)
Throughput for all traffic flows with different background
data rates
Simulation results
• Throughput
)
RT2
(kbps
Throughput
AD-RT1
BET
Background data rate (kbps)
Throughput for all traffic flows with different background
data rates
Simulation results
(ms
)
Delay
• Delay
Background data rate (kbps)
Simulation results
Jitter
• Jitter
SBT data rate (kbps)
Contents
• Introduction
• MPLS & Traffic Engineering
• Thesis Statement
• Preemption in MPLS (MNS-2)
• Proposed Preemption Technique
• Simulation Results
• Conclusion and Future Work
Conclusion and future work
 Allocating a dedicated bandwidth for traffic allows
it to have stable behavior in terms of the throughput.
 Better performance (jitter, delay) is achieved when
serving the preempted traffic at lower QoS level.
 Better network management can be achieved since
the consumed resources in the network are known.
 Better network bandwidth utilization is achieved.
Conclusion and future work
 When a real-time traffic can not get the
minimum requirement of recourses, Should it be
served at:
1- Simple Best-effort level!
2- High best-effort level!
3- blocked!
 A comparison study with different real-time
applications to determine the criteria for each
application is required.