February 20, 2017 Md.

Arafat Nazmul
IP Transmission Network Planning

Layer 2 Network Design

 OSI Model

Open System Interconnection

7. Application Layer
6. Presentation Layer
5. Session Layer
4. Transport Layer
3. Network Layer
2. Data Link Layer
1. Physical Layer
 OSI Model

7. Application Layer 7. Application Layer

>> Software
Interact people to another
Yahoo, Google, Web browser, YouTube
 OSI Model

6. Presentation Layer

>> What is the format of Data 6. Presentation Layer

Video : mp4, mp3, avi
Picture : jpeg, png
Webpage : HTMP

>> Encryption & Decryption

>> Compression & Decompression

 OSI Model

5. Session Layer

>> Create and manage the session via Port No.

5. Session Layer
 OSI Model

4. Transport Layer

>> End to end data delivery from source to destination.

>> TCP/ UDP
>> Flow Control
>> Segmentation/ Fragmentation 4. Transport Layer
>>PC to Destination >> Destination to PC

>>Sequencing
>> Error Correction

TCP UDP
Transmission Control Protocol User Datagram Protocol
Send ACK Send no ACK
Connection Oriented Connection less
Faster Very fast
 OSI Model

3. Network Layer

>> IP Address Tagged:

Source Add + Destination Add

3. Network Layer
 OSI Model

2. Data Link Layer

>> MAC Address Tagged:

Source MAC + Destination MAC

2. Data Link Layer

 OSI Model

1. Physical Layer

>> Physical connection : UTP/Fiber

1. Physical Layer
Network Design - Review

A good network design is modular

and hierarchical, with a clear
separation of functions:
 Core: Resilient, few changes, few
features, high bandwidth, CPU power
 Distribution: Aggregation, redundancy
 Access: Port density, affordability,
security features, many adds, moves
and changes
Network Design - Simple ISP1

Network Border

Core

Distribution

Access
Ethernet Functions

 Source and Destination identification

 MAC addresses

 Detect and avoid frame collisions

 Listen and wait for channel to be available
 If collision occurs, wait a random period before retrying
 This is called CASMA-CD: Carrier Sense Multiple Access with
Collision Detection
Ethernet Frame

SFD = Start of Frame Delimiter

DA = Destination Address
SA = Source Address
CRC = Cyclick Redundancy Check
 Hub

 Receives a frame on one port and sends it out every other port, always.

 Collision domain is not reduced

 Traffic ends up in places where it’s not needed

Hub

Hub

A frame sent by one node is always sent to every other node.

Hubs are also called “repeaters” because they just “repeat”
what they hear.
 Switch

Learns the location of each node by looking at the source address of

each incoming frame, and builds a forwarding table

Forwards each incoming frame to the port where the destination

node is
 Reduces the collision domain
 Makes more efficient use of the wire
 Nodes don’t waste time checking frames not destined to them
 Switch

Forwarding Table Switch

Address Port
AAAAAAAAAAAA 1

BBBBBBBBBBBB 5

B
A
Switches and Broadcast

 A switch broadcasts some frames:

 When the destination address is not found in the table
 When the frame is destined to the broadcast address (FF:FF:FF:FF:FF:FF)
 When the frame is destined to a multicast Ethernet address

 So, switches do not reduce the broadcast domain!

 Switch vs. Router

 Routers
more or less do with IP packets what switches do with
Ethernet frames
A router looks at the IP packet destination and checks its routing table to
decide where to forward the packet
 Some differences:
IP packets travel inside Ethernet frames
IP networks can be logically segmented into subnets
Switches do not usually know about IP, they only deal with Ethernet frames
 Switch vs. Router

 Routers do not forward Ethernet broadcasts. So:

 Switches reduce the collision domain
 Routers reduce the broadcast domain

 Thisbecomes really important when trying to design hierarchical,

scalable networks that can grow sustainably
 Traffic Domains

Router

Switch Switch

Hub Hub Hub Hub

Broadcast Domain Collision Domain  Try to eliminate collision domains

 Try to keep your broadcast domain limited to no
more than 250 simultaneously connected hosts
 Segment your network using routers
 Virtual LANs (VLANs)

 Allow us to split switches into separate (virtual) switches

 Only members of a VLAN can see that VLAN’s traffic

 Inter-vlan traffic must go through a router

 Local VLANs

 2 VLANs or more within a single switch

 Edge ports, where end nodes are connected, are configured as members
of a VLAN

 The switch behaves as several virtual switches, sending traffic only within
VLAN members Switch
VLAN X VLAN Y
Edge ports

VLAN X nodes VLAN Y nodes

 Local VLANs

 Two switches can exchange traffic from one or more VLANs

 Inter-switch links are configured as trunks, carrying frames from all or a

subset of a switch’s VLANs

 Each frame carries a tag that identifies which VLAN it belongs to

 VLAN Trunking Protocol

In Ethernet, the switch has two methods of identifying the VLAN that a frame
belongs to:

 ISL – InterSwitch Link (Cisco proprietary)

 IEEE 802.1Q (standards-based)

 Also known as dot1q
Redundancy

Link Aggregation
 Link Aggregation

 Also known as port bundling, link bundling

 You can use multiple links in parallel as a single, logical link

 For increased capacity
 For redundancy (fault tolerance)

 LACP (Link Aggregation Control Protocol) is a standardized method of

negotiating these bundled links between switches
 Link Aggregation

 Two switches connected via multiple links will send LACPDU packets,
identifying themselves and the port capabilities

 They will then automatically build the logical aggregated links, and then
pass traffic.

 Switche ports can be configured as active or passive

 LACP Operation

 Switches A and B are connected to each other using two sets of Fast
Ethernet ports

 LACP is enabled and the ports are turned on

 Switches start sending LACPDUs, then negotiate how to set up the

aggregation

100 Mbps

Switch A Switch B
100 Mbps
LACPDUs
 LACP Operation

 The result is an aggregated 200 Mbps logical link

 The link is also fault tolerant: If one of the member links fail, LACP will
automatically take that link off the bundle, and keep sending traffic over
the remaining link

100 Mbps

Switch A Switch B
100 Mbps

200 Mbps logical link

 Distributing Traffic in Bundled Links

• Bundled links distribute frames using a hashing algorithm, based on:

– Source and/or Destination MAC address
– Source and/or Destination IP address
– Source and/or Destination Port numbers
• This can lead to unbalanced use of the links, depending on the nature of
the traffic
• Always choose the load-balancing method that provides the most
distribution
Redundancy

Spanning Tree Protocol

 Switching Loop

Switch A Switch B
 Node1 sends a
broadcast frame (e.g.
an ARP request)
Swtich C

Node 1
 Switching Loop

Switch A Switch B
• Switches A, B and C
broadcast node 1’s
frame out every port
Swtich C
except received port

The broadcasts are amplified, creating

Node 1 a broadcast storm
 Spanning Tree Protocol

Switch A Switch B
 STP stop the loops which occurs
when you have multiple links
between switches.
Swtich C
 STP stops avoiding Broadcast
Storms, Multiple Frame Copies.

 STP is open standard (IEEE 802.1D)

Node 1
 Switching Loop
How STP Works ?

Switch A Switch B

 Selecting the Root Bridge

 Selecting the Root Port

Swtich C
 Selecting Designated Port & Non Designated Port

Node 1
 Switching Loop
Selecting The Root Bridge
Priority : 32768
MAC : 0000:0000:0002

Swtich C

 The Bridge with the Best (Lowest) Bridge ID

 Bridge ID = Priority + MAC address of the Switch

Switch A Switch B  Out of all the switched in the network, one is

Priority : 32768 elected as a root bridge that becomes the focal
Priority : 32768 MAC : 0000:0000:0001
MAC : 0000:0000:0003 point in the network.

BPDU: Bridge Protocol Data Unit

 Switching Loop
Selecting The Root Port
Priority : 32768
MAC : 0000:0000:0002
 Shortest path to the Root Bridge
Swtich C RP
DP
 Every Non-Root Bridge looks the best way to go
Root-Bridge

BLK DP

RP  Least cost (Speed)

Switch A Switch B  The Lowest forwarding Switch ID
DP
Priority : 32768  Lowest Physical Port Number
Priority : 32768 MAC : 0000:0000:0001
MAC : 0000:0000:0003

BPDU: Bridge Protocol Data Unit

 Switching Loop
Selecting The Designated Port & Non Designated Port
Priority : 32768
MAC : 0000:0000:0002

DP
Swtich C RP  Least cost (Speed)
 The Lowest forwarding Switch ID
 Lowest Physical Port Number

BLK DP

RP
Switch A Switch B
DP
Priority : 32768
Priority : 32768 MAC : 0000:0000:0001
MAC : 0000:0000:0003

BPDU: Bridge Protocol Data Unit

 Switching Loop
STP Port States

Switch A Switch B

 Blocking 20 sec

 Listening 15 sec
Swtich C
 Learning 15 sec

 Forwarding

 Disable
Node 1
Redundancy

Ethernet Ring Protection Switching (ERPS)

ITU-T G.8032
 Redundancy
G.8032 Objectives and Principles
 Use of standard 802 MAC and OAM frames around the ring. Uses
standard 802.1Q (and amended Q bridges), but with xSTP disabled.
 Ring nodes supports standard FDB MAC learning, forwarding, flush
behavior and port blocking/unblocking mechanisms.
 Prevents loops within the ring by blocking one of the links (either a
pre-determined link or a failed link).
 Monitoring of the ETH layer for discovery and identification of Signal
Failure (SF) conditions.
 Protection and recovery switching within 50 ms for typical rings.
 Total communication for the protection mechanism should consume
a very small percentage of total available bandwidth.
 Redundancy
G.8032 Terms and Concepts

 Ring Protection Link (RPL) – Link designated by mechanism that is blocked during Idle
state to prevent loop on Bridged ring
 RPL Owner – Node connected to RPL that blocks traffic on RPL during Idle state and
unblocks during Protected state
 Link Monitoring – Links of ring are monitored using standard ETH CC OAM messages
(CFM)
 Signal Fail (SF) – Signal Fail is declared when ETH trail signal fail condition is detected
 No Request (NR) – No Request is declared when there are no outstanding conditions
(e.g., SF, etc.) on the node
 Ring APS (R-APS) Messages – Protocol messages defined in Y.1731 and G.8032
 Redundancy
G.8032 Timers
G.8032 specifies the use of different timers to avoid race conditions and unnecessary switching operations

 WTR (Wait to Restore) Timer – Used by the RPL Owner to verify that the ring has stabilized before
blocking the RPL after SF Recovery

 The WTR timer can be configured by the operator. The default time interval is 5 minutes; the time
interval ranges from 1 to 12 minutes

 Hold-off Timers – Used by underlying ETH layer to filter out intermittent link faults
Faults will only be reported to the ring protection mechanism if this timer expires

 The default time interval is 0 seconds; the time interval ranges from 0 to 10 seconds. Faults are
reported to the ring protection Mechanism only if this timer expires.
 Ring Idle State
ETH-CC ETH-CC
A. Physical topology has all nodes connected in a ring ETH-CC
RPL RPL
Owner
ETH-CC
B. ERP guarantees lack of loop by blocking the RPL (link between 6

ETH-CC
ETH-CC
& 1 in figure)

ETH-CC

ETH-CC
C. Logical topology has all nodes connected without a loop.
D. Each link is monitored by its two adjacent nodes using ETH CC ETH-CC ETH-CC
OAM messages ETH-CC ETH-CC
E. Signal Failure as defined in Y.1731, is trigger to ring protection
 Loss of Continuity 2 1 6
RPL
 Server layer failure (e.g. Phy Link Down)
3 4 5
Physical topology

2 1 6

3 4 5

Logical topology
 Protection Switching  Link Failure

RPL RPL
Owner
A. Link/node failure is detected by the nodes adjacent to the
failure.

R-APS(SF)
B. The nodes adjacent to the failure, block the failed link and R-APS(SF) R-APS(SF)
report this failure to the ring using R-APS (SF) message
C. R-APS (SF) message triggers R-APS(SF)

 RPL Owner unblocks the RPL

2 1 6 2 1 6
 All nodes perform FDB flushing RPL RPL
D. Ring is in protection state 3 4 5 3 4 5
E. All nodes remain connected in the logical topology. Physical topology

2 1 6 2 1 6

3 4 5 3 4 5
Logical topology
 Protection Switching  Failure Recovery
R-APS(NR, RB)

RPL RPL
A. When the failed link recovers, the traffic is kept blocked Owner

R-APS(NR,R-APS(NR)
on the nodes adjacent to the recovered link
B. The nodes adjacent to the recovered link transmit R-

RB)
APS(NR) message indicating they have no local request R-APS(NR) R-APS(NR)
present
C. When the RPL Owner receives R-APS(NR) message it R-APS(NR)
Starts WTR timer
2 1 6 2 1 6
D. Once WTR timer expires, RPL Owner blocks RPL and RPL RPL
transmits R-APS (NR, RB) message
3 4 5 3 4 5
E. Nodes receiving the message – perform a FDB Flush
and unblock their previously blocked ports Physical topology

F. Ring is now returned to Idle state 2 1 6 2 1 6

3 4 5 3 4 5
Logical topology
Topology

KHJ

ER
PAN P SS
ub
R ing
-2 BAN
MIR
SHA
ICC
ERPS Sub Ring-3

NILU
DHN UTT
GUL-1

ERPS Ring-1
ERPS Sub Ring-4
Master

MOT NMC
BUET

ERPS Sub Ring-5

ERPS Master Ring rpl owner
ERPS Sub Ring-2 rpl owner
MOG
ERPS Sub Ring-3 rpl owner

ERPS Sub Ring-4 rpl owner

ERPS Sub Ring-5 rpl owner

ERPS Sub Ring-6
ERPS Sub Ring-6 rpl owner

KHL
Questions & Discussion