IT436 Cloud MasterRevision Platform
DashboardMCQ PracticeAnswer KeyFlashcardsExam Night
Made by Youssef Mohamed

IT436 Cloud Master — Interactive revision platform for Cloud Computing students.

Made by Youssef Mohamed

Week 10

Cloud Networking

Progress0%
Topics
Data Center & Scatter-GatherTraffic CharacteristicsNetwork ProblemsPhysical LayoutsFabric Extender (FEX)Networking Implications & Cloud Traffic StatsTree vs Clos & Traffic Directions
All Weeks
W1 Cloud BasicsW2 Physical Layer — ComputeW3 Physical Layer — StorageW4 FC SANW5 IP SANW6 FCoEW10 Cloud NetworkingW12 Control Layer
Dashboard/Week 10
7 topicsAdvanced

Cloud Networking

Data-center network design: traffic patterns, congestion problems, physical layouts, and modern fabrics.

Data Center & Scatter-Gather

Data CenterScatter-GatherIntra-DC Focus

Concept

EN

Technical Definition

A data center is a facility housing critical computing resources (servers, storage, networks, power, cooling). This course focuses on networking within the data center. The scatter-gather pattern (e.g. web search) distributes one request to many servers then aggregates responses — causing synchronized bursts.

ع

شرح مبسّط

مركز البيانات منشأة لاستضافة وإدارة موارد الحوسبة الحرجة (سيرفرات، تخزين، شبكات، تبريد، طاقة). التركيز هنا على الشبكة داخل مركز البيانات وليس شبكات المزود.

💡

Real-World Analogy

A teacher asking 30 students a question at once (scatter) and all answering simultaneously (gather) — a flood at the teacher's desk.

Visual Diagram

Spine 1
Spine 2
Spine 3
⇅ full mesh ⇅
Leaf 1
Leaf 2
Leaf 3
Leaf 4

Every leaf connects to every spine → many equal-cost paths (great for East-West traffic)

Important Terms

Data Center

Facility housing servers, storage, networks, power, and cooling.

Example

Cloud provider's physical infrastructure.

Scatter-Gather

Distribute a request to many nodes, then aggregate responses.

Example

A search query fanned out to many index servers.

Exam Focus

🔥

Remember for Exam

🎯MCQ Hints
  • Scatter-gather causes synchronized bursts → leads to TCP Incast.
  • ~76% of cloud DC traffic is East-West (server-to-server within DC).
📖Key Definitions
  • Scatter-Gather: one request distributed to many servers, responses aggregated.
⚠️Common Mistakes
  • Not linking scatter-gather to TCP Incast congestion.

Traffic Characteristics

VolumeLocalityConcurrent FlowsArrival RateFlow Size

Concept

EN

Technical Definition

Data-center traffic is described by volume, locality (how local the destination is), number of concurrent flows, arrival rate, and flow size (mice vs elephant flows).

ع

شرح مبسّط

تتميز حركة مركز البيانات بعدة خصائص: الحجم، الموقعية (هل الوجهة قريبة؟)، عدد التدفقات المتزامنة، معدل الوصول، وحجم التدفق.

💡

Real-World Analogy

Like describing road traffic: how much, where it's going, how many cars at once, how fast they arrive, and trip length.

Important Terms

Volume

Total amount of traffic.

Example

Terabytes per second across the fabric.

Locality

How close source and destination are (rack, cluster, DC, inter-DC).

Example

Rack 12.9%, Cluster 57.5%, DC 11.9%, Inter-DC 17.7% (typical distribution).

Concurrent Flows

Number of simultaneous connections.

Example

Thousands of TCP flows at once.

Arrival Rate

How fast new flows/packets arrive.

Example

Bursty arrivals during scatter-gather.

Flow Size

Amount of data per flow (mice vs elephant).

Example

Small RPCs (mice) vs big transfers (elephants).

Comparison Tables

Traffic Locality (Typical %)

ScopePercentage
Rack12.9%
Cluster57.5%
Data Center11.9%
Inter-DC17.7%

Exam Focus

🔥

Remember for Exam

🎯MCQ Hints
  • Know all five: volume, locality, concurrent flows, arrival rate, flow size.
  • Most traffic is cluster-local (~57.5%).
  • Mice = small flows; Elephant = large flows.
📖Key Definitions
  • Flow size: data volume of a single flow (mice = small, elephant = large).
⚠️Common Mistakes
  • Forgetting locality as a traffic characteristic.

Network Problems

CongestionTCP IncastIsolation

Concept

EN

Technical Definition

Key problems: congestion (oversubscribed links), TCP Incast (throughput collapse when many synchronized senders overflow a switch buffer to one receiver), and isolation (keeping tenants' traffic separated/fair).

ع

شرح مبسّط

أهم مشاكل الشبكة: الازدحام، وTCP Incast (انهيار الإنتاجية عند وصول ردود كثيرة متزامنة لمنفذ واحد)، ومشكلة العزل بين المستأجرين.

💡

Real-World Analogy

TCP Incast = everyone rushing through one door at once, jamming it so nobody gets through efficiently.

Important Terms

Congestion

Demand exceeds link/buffer capacity, causing delay/drops.

Example

Oversubscribed uplink during peak.

TCP Incast

Throughput collapse from many synchronized senders to one receiver.

Example

Scatter-gather overflowing a ToR switch buffer.

Isolation

Separating/protecting tenant traffic for fairness & security.

Example

One tenant's burst not harming another.

Exam Focus

🔥

Remember for Exam

🎯MCQ Hints
  • TCP Incast = many-to-one synchronized bursts → buffer overflow → throughput collapse.
📖Key Definitions
  • TCP Incast: throughput collapse caused by synchronized many-to-one traffic.
⚠️Common Mistakes
  • Describing TCP Incast as one-to-many — it's many-to-one.

Physical Layouts

ToREoRFabric ExtenderAggregation Switch

Concept

EN

Technical Definition

Cabling layouts: ToR (Top-of-Rack) places a switch in each rack (short cables, more switches); EoR (End-of-Row) centralizes switching at the row's end (longer cables, fewer switches). A fabric extender acts as a remote line card; an aggregation switch concentrates traffic from access switches.

ع

شرح مبسّط

تخطيطات الكابلات: ToR (سويتش أعلى كل رف) يقلل الكابلات الطويلة، وEoR (سويتش نهاية الصف) يركّز الإدارة. Fabric Extender يمدد السويتش، وAggregation Switch يجمع عدة سويتشات.

💡

Real-World Analogy

ToR = a mailbox on every floor; EoR = one mailroom per building wing.

Visual Diagram

Core / Aggregation
↓
Rack A · ToR
Server
Server
Server
Rack B · ToR
Server
Server
Server
Rack C · ToR
Server
Server
Server

Comparison Tables

ToR vs EoR

AspectToR (Top-of-Rack)EoR (End-of-Row)
Switch locationIn each rackAt end of the row
CablingShort, within rackLonger, across row
Switch countMore switchesFewer switches
ManagementMore devices to manageCentralized

Exam Focus

🔥

Remember for Exam

🎯MCQ Hints
  • ToR = switch per rack (short cables, more switches).
  • EoR = switch per row (longer cables, fewer/centralized switches).
  • Why ToR: east-west traffic, copper in-rack, modular per-rack, unified fabric ready.
  • Why not ToR: more switches, power, STP instances, port utilization.
  • Why EoR: fewer switches, fewer STP instances, Layer 1 rack connections.
  • Why not EoR: expensive bulky copper, cable management challenges.
📖Key Definitions
  • ToR: Top-of-Rack switching with a switch in every rack.
⚠️Common Mistakes
  • Swapping ToR and EoR cabling/switch-count tradeoffs.

Fabric Extender (FEX)

Remote Line CardEoR ManagementToR Cabling Benefits

Concept

EN

Technical Definition

A Fabric Extender acts as a remote line card, extending an aggregation switch into the rack (copper stays in-rack) while keeping centralized management. It combines ToR benefits (short cabling, modular per-rack) with EoR benefits (fewer switches to manage, fewer aggregation ports, fewer STP instances).

ع

شرح مبسّط

Fabric Extender يعمل كـ line card بعيد — يمدد switch aggregation إلى داخل الرack (copper in-rack) مع إدارة مركزية. يجمع مزايا ToR (كابلات قصيرة) مع EoR (سويتشات أقل للإدارة).

💡

Real-World Analogy

Like a remote checkout counter (FEX in each rack) controlled by one central office (aggregation switch).

Important Terms

Fabric Extender

Remote I/O module extending a parent switch into racks.

Example

Cisco FEX acting as remote line cards for Nexus.

Parent Switch

Aggregation switch that manages FEX modules.

Example

Core/aggregation switch controlling rack FEXs.

Exam Focus

🔥

Remember for Exam

🎯MCQ Hints
  • FEX = remote line card; fewer switches to manage + in-rack copper.
  • Combines ToR cabling benefits with EoR centralized management.
📖Key Definitions
  • Fabric Extender: extends switch fabric into racks as remote line cards.
⚠️Common Mistakes
  • Treating FEX as a fully independent switch — it's managed by a parent.

Networking Implications & Cloud Traffic Stats

76% East-West25% Annual GrowthVirtualization ImpactTight I/O Deadlines

Concept

EN

Technical Definition

Data-center traffic is massive and ~76% is East-West (server-to-server within the cloud DC). Traffic grows ~25% annually. Server virtualization increases East-West traffic. Implications: tight network I/O deadlines, TCP Incast congestion, need for tenant/application isolation, and difficulty of centralized per-flow control.

ع

شرح مبسّط

حركة مركز البيانات كبيرة و76% منها East-West (بين السيرفرات). النمو ~25% سنوياً. الافتراضية تزيد East-West traffic. التحديات: deadlines ضيقة للـ I/O، TCP Incast، والحاجة لعزل بين التطبيقات.

💡

Real-World Analogy

Most traffic is coworkers talking to each other (East-West), not visitors coming in (North-South).

Important Terms

East-West Dominance

~76% of cloud DC traffic is server-to-server.

Example

VM-to-VM, distributed databases, scatter-gather.

Tight I/O Deadline

DC apps expect low-latency network responses.

Example

Microservices RPCs with strict latency SLAs.

Exam Focus

🔥

Remember for Exam

🎯MCQ Hints
  • 76% of cloud DC traffic is East-West.
  • Virtualization increases East-West traffic.
  • DC traffic grows ~25% per year (Cisco Global Cloud Index).
📖Key Definitions
  • East-West traffic: server-to-server communication within the data center.
⚠️Common Mistakes
  • Underestimating East-West traffic — it dominates cloud DCs.

Tree vs Clos & Traffic Directions

Tree NetworkClos NetworkEast-WestNorth-South

Concept

EN

Technical Definition

Tree networks connect ToR switches to aggregation switches but bottleneck at the top (expensive, hard to scale, congestion). Clos (spine-leaf) uses many small switches with equal-cost paths and high bisection bandwidth. East-West = server-to-server within DC; North-South = in/out of the data center.

ع

شرح مبسّط

شبكة Tree تستخدم Aggregation switches لكنها تعاني من اختناق وتكلفة N². Clos (Spine-Leaf) يحل ذلك بمسارات متعددة. East-West = بين السيرفرات داخل DC (~76% من حركة السحابة). North-South = داخل/خارج DC.

💡

Real-World Analogy

Tree = one narrow highway to downtown (jams); Clos = a grid of equal roads.

Visual Diagram

Spine 1
Spine 2
Spine 3
⇅ full mesh ⇅
Leaf 1
Leaf 2
Leaf 3
Leaf 4

Every leaf connects to every spine → many equal-cost paths (great for East-West traffic)

Comparison Tables

Tree vs Clos Network

AspectTreeClos (Spine-Leaf)
PathsSingle/limitedMany equal-cost paths
BottleneckAt aggregation/coreAvoided (scale-out)
CostExpensive large switchesMany cheap small switches
ScalabilityPoor (N² problem)High bisection bandwidth
Best forTraditional North-SouthEast-West heavy workloads

East-West vs North-South Traffic

DirectionMeaningExample
East-WestServer-to-server inside the DCVM-to-VM, distributed compute
North-SouthIn/out of the data centerClient request from the internet

Exam Focus

🔥

Remember for Exam

🎯MCQ Hints
  • Clos / spine-leaf = many equal paths → great for East-West traffic.
  • East-West = inside DC; North-South = in/out of DC.
  • Tree networks bottleneck at the core.
📖Key Definitions
  • Clos network: multi-stage spine-leaf fabric with high bisection bandwidth.
⚠️Common Mistakes
  • Calling East-West traffic 'in and out of the data center' — that's North-South.
← Week 6: FCoEWeek 12: Control Layer