Twitter Timeline Design

The Challenge

Building a timeline system that can:

Key Constraints

Fan-Out Approaches

Approach 1: Fan-out on Write (Push Model)

When a user tweets, push the tweet to all followers' timelines.

Architecture:

1User tweets → Write to DB → Fan-out service → Update N follower timelines

Pros:

Cons:

Example:

Approach 2: Fan-out on Read (Pull Model)

When a user loads timeline, fetch tweets from all followees.

Architecture:

1User requests timeline → Fetch tweets from N followees → Merge & sort → Return

Pros:

Cons:

Example:

Hybrid Approach (Best Solution)

Use fan-out on write for regular users and fan-out on read for celebrities.

Architecture

1┌─────────────┐
2│ User Tweets │
3└──────┬──────┘
4       │
5       ▼
6┌─────────────────┐
7│  Tweet Service  │
8└──────┬──────────┘
9       │
10       ├─────────────────┬─────────────────┐
11       ▼                 ▼                 ▼
12┌─────────────┐  ┌─────────────┐  ┌─────────────┐
13│ Regular User│  │  Celebrity  │  │   Timeline  │
14│ Fan-out     │  │  Fan-out    │  │   Merge     │
15│ (Push)      │  │  (Pull)     │  │   Service   │
16└─────────────┘  └─────────────┘  └─────────────┘
17       │                 │                 │
18       └─────────────────┴─────────────────┘
19                         │
20                         ▼
21                  ┌──────────────┐
22                  │   Timeline   │
23                  │   Cache      │
24                  └──────────────┘

Implementation Strategy

- Tweet → Push to follower timelines

- Store in Redis cache

- Tweet → Store in DB only

- On timeline load → Fetch celebrity tweets separately

- Sort by timestamp

- Apply ranking algorithm

- Return top N tweets

Database Schema

1-- Tweets table
2CREATE TABLE tweets (
3  id BIGINT PRIMARY KEY,
4  user_id BIGINT,
5  content TEXT,
6  created_at TIMESTAMP
7);
8 
9-- Follows table
10CREATE TABLE follows (
11  follower_id BIGINT,
12  followee_id BIGINT,
13  PRIMARY KEY (follower_id, followee_id)
14);
15 
16-- Timeline cache (Redis)
17-- Key: timeline:{user_id}
18-- Value: List of tweet IDs (sorted by timestamp)

Key Optimizations

1. Caching Strategy

2. Ranking Algorithm

Not just chronological! Consider:

3. Pagination

4. Real-time Updates

Scaling Considerations

Horizontal Scaling

Monitoring

Trade-offs

| Approach | Write Latency | Read Latency | Storage | Complexity |

|----------|---------------|--------------|---------|------------|

| Fan-out Write | High | Low | High | Medium |

| Fan-out Read | Low | High | Low | High |

| Hybrid | Medium | Low | Medium | Very High |

The hybrid approach balances performance with complexity, which is why it's used by Twitter, Facebook, and other social platforms.