Sharding — Animate | AI Wisdom

← Back to topic
🎞 ANIMATE · Mode 3 of 5 · +100 XP— Consistent Hashing
📋Ring Setup
⚙️Virtual Nodes
🔄Route Key
✅Add Server
🎯 Next: Test →
Step 1 / 4
Speed
Create the hash ring
Place servers on a circular ring using MD5 hash positions💡 Analogy
Imagine a clock face. Servers are placed at random positions based on their names.
Hash each server name to get a number (0 to 2^32-1). Place that number on a circular ring.
Console Output
> hash("db1") = 1,432,891,204
> hash("db2") = 2,847,103,882
> hash("db3") = 673,021,556
> Ring: db3(0.15) → db1(0.33) → db2(0.66)
●consistent-hashing.ts
Create the hash ringL6–11
1import crypto from 'crypto';
2 
3class ConsistentHashRing {
4  private ring = new Map<number, string>(); // hash → server
5  private virtualNodes = 150;                // Replicas per server
6 
7  addServer(server: string) {
8    for (let i = 0; i < this.virtualNodes; i++) {
9      const hash = this.hash(`${server}:vnode${i}`);
10      this.ring.set(hash, server);
11    }
12  }
13 
14  getServer(key: string): string {
15    const keyHash = this.hash(key);
16    const sorted = [...this.ring.keys()].sort((a,b) => a-b);
17 
18    // Find first ring position >= keyHash
19    const pos = sorted.find(h => h >= keyHash) ?? sorted[0];
20    return this.ring.get(pos)!;
21  }
22 
23  private hash(input: string): number {
24    const h = crypto.createHash('md5').update(input).digest('hex');
25    return parseInt(h.substring(0,8), 16); // 32-bit hash
26  }
27}
28 
29const ring = new ConsistentHashRing();
30ring.addServer('db1'); ring.addServer('db2'); ring.addServer('db3');
31ring.getServer('user:1234'); // → "db2"
31 lines
Timeline — 4 Steps
Create the hash ring
Place servers on a circular ring using MD5 hash positions💡 Analogy
Imagine a clock face. Servers are placed at random positions based on their names.
Hash each server name to get a number (0 to 2^32-1). Place that number on a circular ring.
Console Output
> hash("db1") = 1,432,891,204
> hash("db2") = 2,847,103,882
> hash("db3") = 673,021,556
> Ring: db3(0.15) → db1(0.33) → db2(0.66)

Was this page helpful?

Discussion

Place servers on a circular ring using MD5 hash positions

💡 Analogy

Imagine a clock face. Servers are placed at random positions based on their names.

Hash each server name to get a number (0 to 2^32-1). Place that number on a circular ring.

Console Output

> hash("db1") = 1,432,891,204
> hash("db2") = 2,847,103,882
> hash("db3") = 673,021,556
> Ring: db3(0.15) → db1(0.33) → db2(0.66)

Place servers on a circular ring using MD5 hash positions

💡 Analogy

Imagine a clock face. Servers are placed at random positions based on their names.

Hash each server name to get a number (0 to 2^32-1). Place that number on a circular ring.

Console Output

> hash("db1") = 1,432,891,204
> hash("db2") = 2,847,103,882
> hash("db3") = 673,021,556
> Ring: db3(0.15) → db1(0.33) → db2(0.66)