Network Layer II Flashcards
(16 cards)
Switching via interconnection network
A switching plane is the part of a router that connects input ports to output ports — it’s the core engine that moves packets inside the router
▪ scaling, using multiple switching “planes” in parallel:
▪ speedup, scaleup via parallelism
▪ Cisco CRS router:
▪ basic unit: 8
switching planes
▪ each plane: 3-stage
interconnection
network
▪ up to 100’s Tbps
switching capacity
Each switching plane uses a 3-stage interconnection network, which allows:
Efficient packet routing from any input port to any output port
High throughput
Non-blocking performance (packets don’t get stuck)
Input port queuing
- If switch fabric slower than input ports combined -> queueing may
occur at input queues - queueing delay and loss due to input buffer overflow!
▪Head-of-the-Line (HOL) blocking: queued datagram at front of queue
prevents others in queue from moving forward
Output port queuing
▪ Buffering required when datagrams
arrive from fabric faster than link
transmission rate. Drop policy: which
datagrams to drop if no free buffers?
▪Scheduling discipline chooses
among queued datagrams for
transmission
Priority scheduling – who
gets best performance,
network neutrality
Packet Scheduling: FCFS
packet scheduling: deciding
which packet to send next on
link
* first come, first served
* priority
* round robin
* weighted fair queueing
Scheduling policies: priority
Priority scheduling:
* arriving traffic classified,
queued by class
- any header fields can be
used for classification
▪ send packet from highest
priority queue that has
buffered packets
* FCFS within priority class
The router always checks the highest priority queue first, and sends a packet from there if it’s not empty. Lower-priority queues are only served when all higher-priority queues are empty.
Scheduling policies: round robin
Round Robin (RR) scheduling:
- arriving traffic classified,
queued by class - any header fields can be
used for classification
▪server cyclically, repeatedly
scans class queues,
sending one complete
packet from each class (if
available) in turn
To treat all classes of traffic fairly, ensuring each class gets a turn to send a packet, regardless of priority.
Scheduling policies: weighted fair queueing
Weighted Fair Queuing (WFQ):
* generalized Round Robin
▪ each class, i, has weight, wi,
and gets weighted amount
of service in each cycle:
wi
wj
j
▪ minimum bandwidth
guarantee (per-traffic-class)
WFQ is an advanced version of Round Robin scheduling that adds fairness with control — it allows different traffic classes to get different amounts of bandwidth, based on their weights.
how much buffering
but too much buffering can increase delays (particularly in home
routers)
- long RTTs: poor performance for real time apps, sluggish TCP response
- recall delay-based congestion control: “keep bottleneck link just full
enough (busy) but no fuller”
IP addressing: introduction
IP address: 32-bit identifier associated with each host or router interface
- interface: connection between host/router and physical link
- router’s typically have multiple interfaces
- host typically has one or two interfaces (e.g., wired Ethernet,
wireless 802.11)
Subnets
A subnet (short for subnetwork) is a smaller part of a larger IP network. It’s used to organize and segment devices within a network.
▪What’s a subnet ?
* device interfaces that can physically reach each other without passing through an intervening router
▪IP addresses have structure:
* subnet part: devices in same subnet
have common high order bits
* host part: remaining low order bits
Recipe for defining subnets:
▪detach each interface from its
host or router, creating
“islands” of isolated networks
▪each isolated network is
called a subnet
Flat addressing
Flat Addressing (Bad for Scaling)
Imagine every device in the world has a completely unique, random address.
Routers would need to remember a separate rule for where to send traffic for every single device.
🧠 That means:
For 4.3 billion possible IPs (2³²), a router would need 4.3 billion entries in its table!
📉 Not scalable.
Classless Inter-Domain Routing (CIDR)
Key ideas:
Flexible division between network and host
addresses
* Get rid of IP classes
* Network prefix can be any size
* A mask is a number that determines the network part and the host
part
How to cope with different MTUs (Maximum
Transmission Units)?
Problem: each network has its own MTU
Maximum datagram size / Maximum Transmission Unit (MTU)
Minimum MTU may not be known for a given path
IP Solution: fragmentation
Split datagrams into pieces when MTU is reduced
where should reassembly happen?
Answer #1: within the network, with no help from end
host B (receiver)
Fragmentation can be harmful
Fragmentation causes inefficient use of resources, same packet
processed twice.
Loss of fragments leads to degraded performance
IPv6: motivation
initial motivation: 32-bit IPv4 address space would be completely allocated
▪ additional motivation:
* speed processing/forwarding: 40-byte fixed length header
* enable different network-layer treatment of “flows”
flows:
This allows routers to identify and handle packets belonging to the same flow (like a video call or stream) in a special or optimized way
Transition from IPv4 to IPv6
- not all routers can be upgraded simultaneously
- no “flag days”
- how will network operate with mixed IPv4 and IPv6 routers?
▪tunneling: IPv6 datagram carried as payload in IPv4 datagram among
IPv4 routers (“packet within a packet”)
* tunneling used extensively in other contexts (4G/5G)
