What is Partition Tolerance (P in CAP)?
Partition tolerance means:
🟢 The system keeps working even if network breaks (communication failure) happen between some nodes.
Simple real-life analogy
You and your friend both use WhatsApp 📶 Your friend’s phone loses internet → network partition
Can WhatsApp still work for you (send/receive)? ✅ Yes → WhatsApp is partition tolerant
In distributed database terms
Suppose you have 2 DB servers (nodes) 🌍 Node A (India) and Node B (US)
→ Network cable cut → Partition (nodes can’t talk)
If the DB is partition tolerant, ✅ Node A still handles requests ✅ Node B still handles requests
They sync later when network comes back
Why Partition Tolerance is essential?
In distributed systems, Network failures WILL happen (it’s not if, it’s when) 🌐 Internet glitches, datacenter issues, cross-region lag
So, real-world distributed systems MUST tolerate partitions (That’s why in CAP — practically every system needs P)
Quick summary table
| Without P | With P |
|---|---|
| System crashes / blocks during network split | System keeps running, syncs later |
Example DBs
| DB | Partition tolerant? | Why? |
|---|---|---|
| Cassandra | ✅ Yes (AP) | Always on, syncs later |
| DynamoDB | ✅ Yes (AP) | Same |
| MongoDB | ✅ Yes (CP or AP mode) | Allows choice |
| MySQL (single node) | ❌ No | Single server → no distributed split |
In short
Partition tolerance means: System keeps running even when network splits happen between nodes → handles failure gracefully.