Write-through and write-back caching in HLD - Architecture Diagram

System Overview - Write-through and write-back caching

This system demonstrates two common caching strategies: write-through and write-back caching. It shows how data is written and read between users, cache, and database to improve performance and data consistency.

Architecture Diagram

User
  |
  v
Load Balancer
  |
  v
API Gateway
  |
  v
+-------------------+
|   Cache Layer      |
|  (Write-through or |
|   Write-back)      |
+-------------------+
  |
  v
Database

Components

User

client

Sends read and write requests to the system

Load Balancer

load_balancer

Distributes incoming requests evenly to API Gateway instances

API Gateway

api_gateway

Receives requests, routes them to cache or database, and handles caching logic

Cache Layer

cache

Stores frequently accessed data to reduce database load; supports write-through or write-back caching

Database

database

Persistent storage for all data

Request Flow - 8 Hops

User → Load Balancer

Load Balancer → API Gateway

API Gateway → Cache Layer

Cache Layer → Database

Database → Cache Layer

Cache Layer → API Gateway

API Gateway → Load Balancer

Load Balancer → User

Failure Scenario

Component Fails:Cache Layer

Impact:Cache misses increase; system falls back to database for reads and writes, increasing latency and load on database

Mitigation:API Gateway bypasses cache and directly queries database; cache rebuilds after recovery

Architecture Quiz - 3 Questions

In write-through caching, when is data written to the database?

A Only when cache is full

B Immediately when data is written to cache

C After a delay or periodically

D Never, only cache stores data

Design Principle

This architecture demonstrates how caching strategies affect data consistency and system performance. Write-through caching ensures data is always consistent by writing to cache and database simultaneously, while write-back caching improves write performance by delaying database writes but risks data loss on failure. The system uses a load balancer and API gateway to manage traffic and caching logic, ensuring scalability and reliability.

Practice

1. Which caching method writes data to both the cache and main storage immediately to ensure data safety?

easy

A. Read-through caching

B. Write-back caching

C. Write-through caching

D. Lazy caching

Solution

1. Step 1: Understand write-through caching behavior

   Write-through caching updates both cache and main storage at the same time to keep data consistent.

2. Step 2: Compare with write-back caching

   Write-back caching delays updating main storage, only writing to cache first for speed.

3. Final Answer:

   Write-through caching -> Option C

4. Quick Check:

   Immediate write to both cache and storage = Write-through caching [OK]

Hint: Immediate write to storage means write-through caching [OK]

Common Mistakes:

• Confusing write-back with write-through
• Thinking write-back writes immediately to storage
• Mixing read-through with write caching

2. Which of the following correctly describes write-back caching?

easy

A. Data is written only to cache first, then to main storage later.

B. Data is written to cache and main storage simultaneously.

C. Data is never written to main storage.

D. Data is written only to main storage, skipping cache.

Solution

1. Step 1: Recall write-back caching process

   Write-back caching writes data to cache first, delaying main storage update for performance.

2. Step 2: Eliminate incorrect options

   The description of data written to cache and main storage simultaneously matches write-through. The claim that data is never written to main storage is false, as main storage is updated eventually. Writing only to main storage skips the cache, which is incorrect.

3. Final Answer:

   Data is written only to cache first, then to main storage later. -> Option A

4. Quick Check:

   Write-back delays main storage write = Data is written only to cache first, then to main storage later. [OK]

Hint: Write-back caches first, writes storage later [OK]

Common Mistakes:

• Choosing simultaneous write as write-back
• Assuming main storage is never updated
• Confusing cache skipping with write-back

3. Consider a system using write-back caching. If a power failure occurs before the cache is flushed to main storage, what is the likely outcome?

medium

A. Some recent data in cache may be lost.

B. All data is safely stored in main storage.

C. Cache and main storage remain fully synchronized.

D. The system automatically switches to write-through caching.

Solution

1. Step 1: Understand write-back caching risk

   Write-back caches data first and delays writing to main storage, so recent changes may only be in cache.

2. Step 2: Analyze power failure impact

   If power fails before cache flush, data in cache not yet saved to main storage is lost.

3. Final Answer:

   Some recent data in cache may be lost. -> Option A

4. Quick Check:

   Power failure before flush = possible data loss in cache [OK]

Hint: Power loss before flush loses cached data [OK]

Common Mistakes:

• Assuming all data is safe despite power failure
• Thinking cache and storage always sync instantly
• Believing system auto-switches caching modes

4. A developer implemented write-back caching but notices data inconsistencies after crashes. Which fix best addresses this issue?

medium

A. Flush cache to storage less frequently.

B. Switch to write-through caching to update storage immediately.

C. Disable caching entirely.

D. Increase cache size to hold more data.

Solution

1. Step 1: Identify cause of inconsistencies

   Write-back caching delays storage updates, causing data loss on crashes.

2. Step 2: Choose solution to ensure data safety

   Switching to write-through caching writes data immediately to storage, preventing inconsistencies.

3. Final Answer:

   Switch to write-through caching to update storage immediately. -> Option B

4. Quick Check:

   Immediate storage update fixes inconsistency = Switch to write-through caching to update storage immediately. [OK]

Hint: Immediate writes fix inconsistency in caching [OK]

Common Mistakes:

• Thinking bigger cache solves data loss
• Disabling cache reduces speed unnecessarily
• Flushing less often worsens data loss

5. You are designing a high-performance system that must balance speed and data reliability. Which caching strategy should you choose and why?
Options:
A) Write-through caching for data safety but slower writes.
B) Hybrid approach: write-back with periodic flush and battery-backed cache.
C) Write-back caching for faster writes but risk of data loss.
D) No caching to avoid complexity.

hard

A. Write-through caching for data safety but slower writes.

B. No caching to avoid complexity.

C. Write-back caching for faster writes but risk of data loss.

D. Hybrid approach: write-back with periodic flush and battery-backed cache.

Solution

1. Step 1: Analyze trade-offs of write-through and write-back

   Write-through is safe but slower; write-back is fast but risks data loss on failure.

2. Step 2: Consider hybrid approach benefits

   Hybrid uses write-back speed plus periodic flush and battery backup to reduce data loss risk.

3. Step 3: Evaluate no caching option

   No caching avoids complexity but sacrifices performance, not ideal for high-performance needs.

4. Final Answer:

   Hybrid approach: write-back with periodic flush and battery-backed cache. -> Option D

5. Quick Check:

   Balance speed and safety with hybrid caching = Hybrid approach: write-back with periodic flush and battery-backed cache. [OK]

Hint: Hybrid caching balances speed and safety best [OK]

Common Mistakes:

• Choosing only write-back ignoring data loss
• Picking no caching sacrificing performance
• Ignoring battery-backed cache benefits