🔹 What are Redo Logs?

Redo Logs are a mechanism used by databases to ensure durability and recoverability.

• They record all changes made to the database so that in case of a crash, the database can redo the changes and recover to a consistent state.
• Conceptually, they are very similar to Write-Ahead Logs (WAL).

---

🔹 How Redo Logs Work

1. Transaction begins → changes are made in memory (buffer/cache).
2. Changes are written to redo log first (sequentially).
3. Actual database files are updated later (asynchronously).

4. On crash recovery:

   • Database reads redo logs.
   • Re-applies all committed changes to bring database back to a consistent state.

---

🔹 Example

Suppose you execute an UPDATE accounts SET balance = balance - 100 WHERE id=1;

Step 1: Append to redo log
[Redo Log] UPDATE accounts SET balance = balance - 100 WHERE id=1;

Step 2: Update in-memory database buffer
[Buffer] balance = balance - 100

Step 3: Flush buffer to disk (DB file)

Crash happens? Recovery reads redo log:
- Re-applies update → ensures transaction is not lost

---

🔹 Key Properties

1. Durability (ACID) ✅

   • Once transaction is written to redo log, it’s considered durable.

2. Crash Recovery 🔄

   • Redo logs allow database to replay committed changes after a crash.

3. Sequential Writes 📄

   • Writing to redo log is sequential → faster than random disk writes.

4. Buffering Optimization

   • Database can flush multiple redo log entries together → improves performance.

---

🔹 Redo Log vs WAL

Feature	Redo Log	WAL (Write-Ahead Log)
Purpose	Recover committed changes	Recover committed changes
Write Order	Log written before data file	Same: log written before data file
Typical Use	MySQL (InnoDB), Oracle	PostgreSQL, SQLite, MongoDB (journal)
Granularity	Usually at transaction level	Transaction or statement level
Storage	Usually fixed-size circular logs	Append-only sequential logs

---

🔹 Usage in Databases

• MySQL (InnoDB) → Uses redo logs to implement durability.
• Oracle → Redo logs are core to transaction recovery.
• PostgreSQL / SQLite → Equivalent concept via WAL.
• MongoDB → Journaling acts like redo logs.

---

💡 Summary:

• Redo Log = sequential log of all changes that lets DB recover committed transactions after a crash.
• It’s basically the implementation of WAL in databases like MySQL/Oracle.

---