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 3

Physical Layer — Storage

Progress0%
Topics
HDD InternalsSSD vs HDDRAID Techniques & LevelsCacheStorage Access MethodsNAS Deployment OptionsStorage VirtualizationLUN, LUN Masking & Storage PoolsMetaLUN & LUN ExpansionTraditional vs Thin/Thick ProvisioningStorage System ArchitectureRAID Write Penalties & Read Operations
All Weeks
W1 Cloud BasicsW2 Physical Layer — ComputeW3 Physical Layer — StorageW4 FC SANW5 IP SANW6 FCoEW10 Cloud NetworkingW12 Control Layer
Dashboard/Week 3
12 topicsIntermediate

Physical Layer — Storage

From spinning platters to RAID, caching, storage access methods, and provisioning strategies.

HDD Internals

PlatterHeadTrackSector

Concept

EN

Technical Definition

A hard disk drive stores data magnetically on rotating platters. A read/write head accesses data organized into concentric tracks, each divided into sectors.

ع

شرح مبسّط

القرص الصلب الميكانيكي يخزن البيانات على أقراص دوارة (Platters) يقرأها رأس (Head). كل قرص مقسم إلى مسارات (Tracks) وكل مسار إلى قطاعات (Sectors).

💡

Real-World Analogy

Like a vinyl record: the platter spins, the needle (head) reads grooves (tracks) split into segments (sectors).

Important Terms

Platter

Rotating magnetic disk that stores data.

Example

A 7200 RPM platter.

Head

Reads/writes data on the platter surface.

Example

One head per platter surface.

Track

Concentric circle on a platter.

Example

Outer tracks hold more sectors.

Sector

Smallest addressable storage unit on a track.

Example

Traditionally 512 bytes.

Exam Focus

🔥

Remember for Exam

🎯MCQ Hints
  • Order from large to small: Platter → Track → Sector.
  • Sector is the smallest addressable unit.
📖Key Definitions
  • Sector: smallest addressable storage unit on a disk.
⚠️Common Mistakes
  • Confusing track (a circle) with sector (a segment of a track).

SSD vs HDD

Concept

EN

Technical Definition

SSDs use flash memory with no moving parts — faster, more durable, lower latency. HDDs use mechanical platters — cheaper per GB, higher capacity but slower.

ع

شرح مبسّط

SSD يستخدم ذاكرة فلاش بدون أجزاء متحركة فيكون أسرع وأكثر تحملاً، بينما HDD أرخص لكل جيجابايت لكنه أبطأ.

💡

Real-World Analogy

SSD is instant memory recall; HDD is fetching a book from a spinning shelf.

Comparison Tables

SSD vs HDD

AspectSSDHDD
Moving partsNoneYes (platters/heads)
SpeedVery fastSlower
LatencyLow (microseconds)Higher (milliseconds)
Cost per GBHigherLower
DurabilityMore shock-resistantSensitive to shock

Exam Focus

🔥

Remember for Exam

🎯MCQ Hints
  • SSD = no moving parts, lower latency.
  • HDD = lower cost per GB.
📖Key Definitions
  • SSD: solid-state drive using flash memory, no mechanical parts.
⚠️Common Mistakes
  • Claiming SSDs are cheaper per GB than HDDs.

RAID Techniques & Levels

StripingMirroringParityRAID 0/1/3/5/6/1+0

Concept

EN

Technical Definition

RAID (Redundant Array of Independent Disks) combines multiple disks for performance and/or fault tolerance using striping, mirroring, and parity.

ع

شرح مبسّط

RAID يدمج عدة أقراص لتحسين الأداء أو الحماية. التقنيات الأساسية: التقسيم (Striping) للأداء، النسخ (Mirroring) للحماية، والتماثل (Parity) لاسترجاع البيانات.

💡

Real-World Analogy

Striping = splitting work among workers (fast); mirroring = a backup twin; parity = a clue card to rebuild lost data.

Visual Diagram

RAID 0StripingNo redundancy
RAID 1Mirroring1 disk fault
RAID 3Stripe + parity disk1 disk fault
RAID 5Distributed parity1 disk fault
RAID 6Double parity2 disk faults
RAID 1+0Mirror + stripeSpeed + safety

Important Terms

Striping

Spreads data across disks for speed (no redundancy).

Example

RAID 0.

Mirroring

Duplicates data on a second disk for redundancy.

Example

RAID 1.

Parity

Stores calculated data to rebuild a failed disk.

Example

RAID 5/6.

Comparison Tables

RAID Levels

LevelTechniqueMin DisksFault ToleranceNote
RAID 0Striping2NoneFastest, no protection
RAID 1Mirroring21 disk50% usable capacity
RAID 3Striping + dedicated parity3+1 diskSingle parity disk; parallel I/O
RAID 5Striping + distributed parity31 diskSingle distributed parity
RAID 6Striping + double parity42 disksDouble parity
RAID 1+0Mirror + stripe41 per mirrorSpeed + redundancy (RAID 10)

Exam Focus

🔥

Remember for Exam

🎯MCQ Hints
  • RAID 0 = no fault tolerance (striping only).
  • RAID 3 = single dedicated parity disk; all disks involved in parallel I/O.
  • RAID 5 tolerates 1 disk failure; RAID 6 tolerates 2.
  • RAID 1+0 (RAID 10) = mirroring + striping combined.
  • RAID 6 write penalty: 6 I/O ops (3 reads + 3 writes) vs RAID 5's 4.
📖Key Definitions
  • Parity: redundant data used to reconstruct a failed disk.
⚠️Common Mistakes
  • Saying RAID 0 provides redundancy — it does not.

Cache

Read Hit / Read MissWrite Through / Write BackLRU / MRUCache WatermarkingCache Mirroring / Cache Vaulting

Concept

EN

Technical Definition

Cache is fast memory holding frequently accessed data. Read hit = data found in cache; read miss = must fetch from disk. Write policies and replacement algorithms manage it.

ع

شرح مبسّط

الكاش ذاكرة سريعة تخزن البيانات المتكررة لتسريع الوصول. عند الطلب: إذا وُجد في الكاش = Read Hit، وإلا = Read Miss.

💡

Real-World Analogy

Cache is your desk's top drawer: things you use often stay close; a 'miss' means walking to the storage room.

Important Terms

Read Hit

Requested data is already in cache.

Example

Fast response, no disk access.

Read Miss

Data not in cache; fetched from disk.

Example

Slower; then cached for next time.

Write Through

Writes go to cache AND disk immediately.

Example

Safer, but slower writes.

Write Back

Writes go to cache first, disk later.

Example

Faster, but risk on power loss.

LRU

Least Recently Used — evicts the oldest-used data.

Example

Drops data not touched for longest time.

MRU

Most Recently Used — evicts the newest-used data.

Example

Useful for streaming/sequential scans.

Idle Flushing

Flushes cache to disk when the system is idle.

Example

Commits data during low activity.

High Watermark Flushing

Flushes when cache utilization hits high watermark (HWM).

Example

Prevents cache from filling completely.

Forced Flushing

Immediately flushes all cache data to storage.

Example

Used under critical cache pressure.

Cache Mirroring

Duplicates cache to protect against controller failure.

Example

Two controllers hold the same write cache.

Cache Vaulting

Flushes cache to persistent storage on power loss.

Example

Battery-backed cache dumped to disk.

Exam Focus

🔥

Remember for Exam

🎯MCQ Hints
  • Write Through = safer (writes to disk immediately); Write Back = faster (risk on power failure).
  • LRU evicts least recently used; MRU evicts most recently used.
  • Cache Vaulting protects write-back cache during power loss.
  • Watermarking: Idle, High watermark (HWM), and Forced flushing manage cache utilization.
📖Key Definitions
  • Write Back: data written to cache first and to disk later.
⚠️Common Mistakes
  • Swapping Write Through and Write Back behavior.

Storage Access Methods

Block-Level (SAN)File-Level (NAS)Object Storage

Concept

EN

Technical Definition

Storage can be accessed at the block level (SAN — raw blocks over a network), file level (NAS — shared file system), or as objects (object storage with metadata and an API).

ع

شرح مبسّط

هناك ثلاث طرق للوصول للتخزين: على مستوى الكتل (SAN)، على مستوى الملفات (NAS)، والتخزين الكائني (Object).

💡

Real-World Analogy

Block = raw bricks you build with; File = an organized filing cabinet; Object = parcels with labels stored in a warehouse.

Comparison Tables

SAN vs NAS vs Object

AspectSAN (Block)NAS (File)Object
Access unitBlocksFilesObjects + metadata
ProtocolFC, iSCSINFS, SMBHTTP/REST API
PerformanceHighestModerateScalable, not low-latency
Best forDatabases, VMsFile sharingBackups, media, cloud

Exam Focus

🔥

Remember for Exam

🎯MCQ Hints
  • SAN = block-level; NAS = file-level; Object = API/metadata-based.
  • Databases prefer block (SAN).
  • Traditional NAS = scale-up; Scale-out NAS = cluster of nodes pooled.
📖Key Definitions
  • Object storage: data stored as objects with metadata, accessed via API.
⚠️Common Mistakes
  • Confusing SAN (block) with NAS (file).

NAS Deployment Options

Traditional NAS (Scale-up)Scale-out NAS

Concept

EN

Technical Definition

NAS deployment options: Traditional (scale-up) NAS upgrades a single system's capacity/performance; Scale-out NAS pools multiple processing and storage nodes in a cluster that works as a single NAS device, scaling without disruption.

ع

شرح مبسّط

NAS يمكن نشره بطريقتين: تقليدي (Scale-up) بتوسيع جهاز واحد، أو Scale-out بتجميع عدة عقد في Cluster يعمل كـ NAS واحد.

💡

Real-World Analogy

Scale-up = making one filing cabinet taller; scale-out = adding more cabinets managed as one library.

Comparison Tables

Traditional vs Scale-out NAS

AspectTraditional (Scale-up)Scale-out
ScalingUpgrade single systemAdd nodes to cluster
ArchitectureSingle file serverCluster of nodes
DisruptionMay require downtimeNon-disruptive expansion

Exam Focus

🔥

Remember for Exam

🎯MCQ Hints
  • Scale-out NAS = multiple nodes pooled as one NAS device.
📖Key Definitions
  • Scale-out NAS: cluster of nodes providing unified NAS capacity and performance.
⚠️Common Mistakes
  • Confusing scale-up (single box) with scale-out (cluster).

Storage Virtualization

Host-BasedArray-BasedNetwork-Based

Concept

EN

Technical Definition

Storage virtualization abstracts physical storage into logical pools. It can be implemented host-based, array-based, or network-based.

ع

شرح مبسّط

افتراضية التخزين تخفي تفاصيل الأقراص الفعلية وتقدمها كمورد موحد. يمكن أن تكون على المضيف، أو على المصفوفة، أو على الشبكة.

💡

Real-World Analogy

Like merging many hard drives so the user sees one big drive.

Important Terms

Host-Based

Virtualization done at the server/host level.

Example

Logical Volume Manager (LVM).

Array-Based

Done within the storage array controller.

Example

Storage array presenting LUNs.

Network-Based

Done in the SAN network (e.g. a switch/appliance).

Example

Virtualization appliance in the fabric.

Exam Focus

🔥

Remember for Exam

🎯MCQ Hints
  • Three locations: host, array, network.
📖Key Definitions
  • Storage virtualization: abstracting physical storage into logical pools.
⚠️Common Mistakes
  • Mixing up where each type is implemented.

LUN, LUN Masking & Storage Pools

LUNLUN MaskingStorage Pool

Concept

EN

Technical Definition

A LUN (Logical Unit Number) is a logical disk presented to a host. LUN masking controls which hosts can see which LUNs (access control). A storage pool aggregates physical disks to carve out LUNs.

ع

شرح مبسّط

LUN هو قرص منطقي يُقدَّم للمضيف من مخزن التخزين. LUN Masking يتحكم بأي مضيف يرى أي LUN. وStorage Pool هو تجميع للأقراص لإنشاء LUNs منها.

💡

Real-World Analogy

Storage pool = a shared water tank; LUNs = labeled bottles filled from it; LUN masking = who's allowed to take which bottle.

Important Terms

LUN

A logical unit (virtual disk) presented to a host.

Example

A 500GB LUN for a database server.

LUN Masking

Restricts which hosts can access specific LUNs.

Example

Only the DB server sees the DB LUN.

Storage Pool

Aggregated physical disks used to create LUNs.

Example

A pool of 20 disks carved into many LUNs.

Exam Focus

🔥

Remember for Exam

🎯MCQ Hints
  • LUN masking = access control at the storage array (which host sees which LUN).
📖Key Definitions
  • LUN: a logical disk unit presented to a host.
⚠️Common Mistakes
  • Confusing LUN masking (array-side access control) with zoning (switch-side).

MetaLUN & LUN Expansion

Concatenated MetaLUNStriped MetaLUN

Concept

EN

Technical Definition

MetaLUN expands LUNs by combining two or more LUNs. Concatenated metaLUN provides additional capacity only (quick expansion, no restriping). Striped metaLUN provides capacity and performance but expansion is slow because data is restriped.

ع

شرح مبسّط

MetaLUN يوسّع LUN بدمج عدة LUNs. Concatenated يضيف سعة فقط، بينما Striped يضيف سعة وأداء لكنه أبطأ لأنه يعيد توزيع البيانات.

💡

Real-World Analogy

Concatenated = taping two notebooks end-to-end; striped = rewriting pages across both for speed.

Comparison Tables

Concatenated vs Striped MetaLUN

AspectConcatenatedStriped
CapacityYesYes
PerformanceNo gainImproved
Expansion speedFast (no restripe)Slow (restripe required)

Exam Focus

🔥

Remember for Exam

🎯MCQ Hints
  • Concatenated = capacity only; Striped = capacity + performance.
📖Key Definitions
  • MetaLUN: method to expand LUNs requiring additional capacity or performance.
⚠️Common Mistakes
  • Thinking concatenated metaLUN improves performance — it does not.

Traditional vs Thin/Thick Provisioning

Concept

EN

Technical Definition

Traditional provisioning (thick) allocates full capacity upfront from a RAID set. Virtual provisioning (thin) allocates on demand from a storage pool as data is written. Thin LUNs can be expanded and the pool rebalanced as drives are added.

ع

شرح مبسّط

Traditional (Thick) provisioning يخصص كل المساحة مقدماً. Virtual (Thin) provisioning يخصص عند الحاجة فقط، مع إمكانية توسيع Thin LUNs وإعادة توازن الـ Storage Pool.

💡

Real-World Analogy

Thick = reserving a full water tank now; Thin = filling the tank only as you use water.

Comparison Tables

Traditional vs Virtual (Thin) Provisioning

AspectTraditional (Thick)Virtual (Thin)
AllocationUpfront from RAID setOn demand from pool
UtilizationLower (unused allocated)Higher (over-provisioning)
ExpansionCreate new LUNExpand thin LUN + pool rebalance
RiskWasted spaceOver-commit / out of space

Thin vs Thick Provisioning

AspectThinThick
AllocationOn demandUpfront (full)
UtilizationHigherLower
RiskOver-commit / out of spaceWasted space
PerformanceSlight overheadPredictable

Exam Focus

🔥

Remember for Exam

🎯MCQ Hints
  • Thin = allocate on demand; Thick = allocate all upfront.
📖Key Definitions
  • Thin provisioning: storage allocated as needed, not upfront.
⚠️Common Mistakes
  • Saying thin provisioning reserves all space at creation.

Storage System Architecture

Front EndBack EndCacheControllers

Concept

EN

Technical Definition

A storage system has a front end connecting to hosts (via SAN/NAS ports), a back end managing physical drives, and controllers with cache in between. The front end handles host I/O; the back end handles RAID, tiering, and drive management.

ع

شرح مبسّط

نظام التخزين له واجهة أمامية (Front End) تتصل بالسيرفرات عبر SAN/NAS، وواجهة خلفية (Back End) تدير الأقراص الفيزيائية، مع كاش وcontrollers في المنتصف.

💡

Real-World Analogy

Front end = store checkout counters; back end = warehouse shelves; cache = items kept at the counter for speed.

Important Terms

Front End

Host-facing ports and protocols (FC, iSCSI, NFS).

Example

FC ports connecting to server HBAs.

Back End

Physical drive enclosure and RAID management.

Example

Disk trays managed by the array controller.

Storage Controller

Brain of the array — handles I/O, cache, RAID.

Example

Dual controllers for high availability.

Exam Focus

🔥

Remember for Exam

🎯MCQ Hints
  • Front end = host connection; back end = physical disks.
📖Key Definitions
  • Front end: the interface through which hosts access storage.
⚠️Common Mistakes
  • Confusing front end (host side) with back end (disk side).

RAID Write Penalties & Read Operations

RAID 1: 2 writesRAID 5: 4 I/O opsRAID 6: 6 I/O opsRead Hit/Miss with Cache

Concept

EN

Technical Definition

Each RAID level has a write penalty: RAID 1 requires two disk writes per write; RAID 5 requires 4 I/O operations (2 reads of old data/parity + 2 writes); RAID 6 requires 6 I/O operations (3 reads + 3 writes with double parity). Cache reduces disk reads on cache hits.

ع

شرح مبسّط

كل RAID level له 'غرامة كتابة': RAID 1 يكتب مرتين، RAID 5 يحتاج 4 عمليات I/O (2 reads + 2 writes)، RAID 6 يحتاج 6 (3 reads + 3 writes). الكاش يقلل قراءة الأقراص عند Read Hit.

💡

Real-World Analogy

Write penalty = extra bookkeeping — mirroring doubles work; parity triples or more.

Comparison Tables

RAID Write Penalties

RAID LevelWrites per Host WriteI/O Operations
RAID 12 disk writes2 writes
RAID 51 host write4 I/O (2 read + 2 write)
RAID 61 host write6 I/O (3 read + 3 write)

Exam Focus

🔥

Remember for Exam

🎯MCQ Hints
  • RAID 1: every write = 2 disk writes.
  • RAID 5 write penalty = 4 I/O ops; RAID 6 = 6 I/O ops.
  • Read Hit = data in cache (fast); Read Miss = fetch from disk.
📖Key Definitions
  • Write penalty: extra I/O operations required due to redundancy mechanism.
⚠️Common Mistakes
  • Thinking RAID 5 and RAID 6 have the same write penalty.
← Week 2: Physical Layer — ComputeWeek 4: FC SAN