Oracle RAC — Complete Guide

Introduction

Oracle RAC — Complete Guide is essential for developers and DBAs building OracleCore Enterprise Oracle Platform — Toolliyo's 96-article Oracle master path covering installation, architecture, SQL*Plus, multitenant PDBs, tablespaces, PL/SQL, RMAN, Flashback, AWR, RAC, Data Guard, OCI, 23ai features, and enterprise OracleCore projects. Every article includes EXPLAIN plans, SGA/PGA internals, transaction flows, and minimum 2 ultra-detailed enterprise database examples (banking RAC, airline reservations, telecom billing, ERP multitenant, healthcare TDE, OCI Autonomous, Data Guard DR).

In Indian IT and product companies (TCS, Infosys, HDFC, Flipkart), interviewers expect oracle rac with real banking transactions, e-commerce scale, deadlock handling, and query tuning — not toy SELECT * demos. This article delivers two mandatory enterprise examples on Airline Reservations.

After this article you will

• Explain Oracle RAC in plain English and in Oracle SQL / instance architecture terms
• Apply oracle rac inside OracleCore Enterprise Oracle Platform (Airline Reservations)
• Compare naive literal SQL vs OracleCore indexed, bind-variable, and AWR-monitored production patterns
• Answer fresher, mid-level, and senior Oracle SQL, PL/SQL, RAC, and DBA interview questions confidently
• Connect this lesson to Article 76 and the 96-article Oracle roadmap

Prerequisites

• Software: Oracle 23ai, SQL Developer or SQL*Plus
• Knowledge: Basic computer literacy
• Previous: Article 74 — Explain Plans — Complete Guide
• Time: 28 min reading + 30–45 min hands-on

Concept deep-dive

Level 1 — Analogy

Oracle RAC on OracleCore teaches Oracle step by step — architecture, PL/SQL, RAC, Data Guard, and enterprise database patterns.

Level 2 — Technical

Oracle RAC powers enterprise databases in OracleCore: normalized schemas, tuned indexes, ACID transactions, AWR monitoring, and secure bind-variable SQL. OracleCore implements Airline Reservations with RAC, Data Guard, and RMAN production patterns.

Level 3 — Query execution flow

[App / .NET / Java / PL/SQL]
       ▼
[Oracle Net → Listener → Service/PDB]
       ▼
[Parse → Optimize (CBO) → Execute]
       ▼
[B-tree indexes / Row locks / Redo log]
       ▼
[AWR · ADDM · RMAN · Data Guard]

Project structure

OracleCore/
├── tablespaces/         ← Datafiles and storage
├── schema/              ← Tables, views, constraints
├── indexes/             ← B-tree and bitmap indexes
├── plsql/               ← Packages and procedures
├── security/            ← Users, roles, TDE
├── ha/                  ← RAC + Data Guard
└── monitoring/          ← AWR · OEM · ADRCI

Step-by-Step Implementation — OracleCore (Airline Reservations)

Follow: design schema → write bind-variable SQL → add indexes → run EXPLAIN PLAN → wrap in transaction → enable AWR → integrate into OracleCore Airline Reservations.

Step 1 — Anti-pattern (literal SQL, no index, full scan)

-- ❌ BAD — literal SQL + full table scan
SELECT * FROM orders WHERE customer_id = ${customerId};
-- Hard-coded literal; no bind variable; no index on customer_id

Step 2 — Production Oracle SQL

-- ✅ PRODUCTION — Oracle RAC on OracleCore (Airline Reservations)
SELECT order_id, order_date, total
FROM orders
WHERE customer_id = :customer_id
ORDER BY order_date DESC
FETCH FIRST 50 ROWS ONLY;
-- Bind variable; index on (customer_id, order_date)

Step 3 — Full script

SELECT sql_id, elapsed_time, executions FROM v$sql
WHERE elapsed_time > 1000000 ORDER BY elapsed_time DESC FETCH FIRST 10 ROWS ONLY;

-- Verify in SQL Developer: EXPLAIN PLAN + AWR top SQL
-- Check ADDM findings after deploy

The problem before Oracle — Oracle RAC

Mission-critical systems need proven ACID, HA, and DBA tooling. OracleCore replaces fragile setups with RAC, Data Guard, RMAN, and enterprise-grade security.

• ❌ Single-instance DB with no DR — hours of downtime on hardware failure
• ❌ Manual backups without RMAN validation — untested restore when crisis hits
• ❌ Full table scans on billion-row tables — billing batch misses SLA
• ❌ Shared SYSDBA credentials — audit failure and security breach risk

OracleCore applies Oracle architecture, indexing, RMAN, and HA patterns from day one.

Database architecture

Oracle RAC in OracleCore module Airline Reservations — category: PERFORMANCE.

AWR, ADDM, explain plans, RAC, ASM, and Data Guard HA.

[App / .NET / Java / PL/SQL]
       ↓
[Oracle Net → Listener → Service/PDB]
       ↓
[Instance: SGA + PGA + Background Processes]
       ↓
[Database Files: Datafiles, Redo, Control]
       ↓
[AWR · ADDM · RMAN · Data Guard]

SQL execution flow

Real-world example 1 — Airtel Telecom Billing at Scale

Domain: Telecom. CDR ingestion generates billions of rows monthly. OracleCore uses interval partitioning, parallel DML, and materialized views for billing summaries.

Architecture

cdr_records partitioned monthly
  local indexes on subscriber_id
  MV refresh FAST ON COMMIT for daily totals
  RMAN incremental backup nightly

Oracle SQL / PL/SQL

CREATE TABLE cdr_records (
  cdr_id NUMBER,
  subscriber_id NUMBER NOT NULL,
  call_date DATE NOT NULL,
  duration_sec NUMBER,
  charge NUMBER(10,2)
) PARTITION BY RANGE (call_date) INTERVAL (NUMTOYMINTERVAL(1,'MONTH'))
(PARTITION p0 VALUES LESS THAN (DATE '2025-01-01'));

Outcome: Billing cycle 8h → 45min; storage compression 3:1 ratio.

Real-world example 2 — AWR-Driven SQL Tuning

Domain: Performance. Checkout API slowed after prodiv. OracleCore uses AWR report, SQL Tuning Advisor, and new composite index on (customer_id, order_date).

Architecture

AWR snapshot every 30 min
  ADDM findings → index recommendations
  SQL Plan Baseline for stable plans
  OEM alerts on top SQL regressions

Oracle SQL / PL/SQL

SELECT * FROM TABLE(DBMS_XPLAN.DISPLAY_CURSOR(NULL, NULL, 'ALLSTATS LAST'));

CREATE INDEX idx_orders_cust_date ON orders (customer_id, order_date) COMPRESS 1;

Outcome: Query elapsed 12s → 80ms; CPU wait eliminated.

DBA & performance tips

• Use bind variables in application SQL — reduces hard parses in shared pool
• Review AWR top SQL and ADDM findings weekly on production
• Test RMAN restore quarterly — backup without tested restore is worthless
• Document RAC/Data Guard runbooks before go-live

When not to use this Oracle pattern for Oracle RAC

• 🔴 Small dev apps — Oracle licensing and ops overhead may exceed benefit
• 🔴 Document-heavy flexible schema — consider NoSQL or JSON-first stores
• 🔴 RAC before exhausting single-instance tuning and Data Guard
• 🔴 Over-partitioning tiny tables — management cost exceeds query benefit

Testing & validation

-- Manual assertion
SELECT COUNT(*) INTO v_count FROM ORACLERAC WHERE status = 'ACTIVE';
-- Assert v_count = expected

Pattern recognition

Lookup by PK → index unique scan. Join heavy → index FK columns. Reporting → materialized views. Money moves → explicit COMMIT. Read scale → Active Data Guard standby. Slow after deploy → AWR top SQL.

Common errors & fixes

🔴 Mistake 1: Literal SQL with concatenated user input
✅ Fix: Use bind variables — prevents SQL injection and reduces hard parses.

🔴 Mistake 2: Missing indexes on WHERE/JOIN columns
✅ Fix: Create B-tree indexes on FK and filter columns used in joins.

🔴 Mistake 3: Long-running transactions holding row locks
✅ Fix: Keep transactions short; COMMIT minimal work units.

🔴 Mistake 4: Ignoring EXPLAIN PLAN and AWR reports
✅ Fix: Run EXPLAIN PLAN on new queries; review AWR top SQL weekly.

Best practices

• 🟢 Use bind variables — never concatenate user input into SQL
• 🟢 Index WHERE and JOIN columns on large Oracle tables
• 🟡 Enable AWR snapshots on every production database from day one
• 🟡 Run EXPLAIN PLAN after schema or data volume changes
• 🔴 Never run money/inventory updates outside explicit transactions
• 🔴 Never deploy without backup strategy and tested restore procedure

Interview questions

Fresher level

Q1: Explain Oracle RAC in a database design interview.
A: Cover schema, indexes, normalization trade-offs, concurrency, security, backup/HA, and monitoring.

Q2: B-tree vs bitmap index in Oracle?
A: B-tree is default for OLTP equality/range; bitmap suits low-cardinality DWH columns.

Q3: What is Oracle RAC?
A: Multiple instances share ASM storage; Clusterware handles node failover for continuous service.

Mid / senior level

Q4: How do you find and fix a slow query?
A: AWR top SQL → EXPLAIN PLAN → missing index? → add index → verify in next AWR.

Q5: Explain deadlock and how to prevent it.
A: Circular lock wait — consistent lock order, shorter transactions, retry in app.

Q6: How do you secure Oracle?
A: Least-privilege roles, no shared SYSDBA in apps, TDE, unified auditing, Oracle Net encryption.

Coding round

Write Oracle SQL for Oracle RAC in OracleCore Airline Reservations: show CREATE script, sample query, explain plan notes, and test assertions.

-- OracleRAC validation
SELECT COUNT(*) AS actual FROM oraclerac WHERE is_active = 1;
-- Assert actual = expected

Summary & next steps

• Article 75: Oracle RAC — Complete Guide
• Module: Module 9: Performance & High Availability · Level: ADVANCED
• Applied to OracleCore — Airline Reservations

Previous: Explain Plans — Complete Guide
Next: Oracle ASM — Complete Guide

Practice: Run today's SQL in SQL Developer with EXPLAIN PLAN — commit with feat(oracle): article-75.

FAQ