Confidentiality
Only authorized identities can read protected data.
Controls: access rules, encryption, masking
CodeBhavyaApply least privilege, prevent injection, plan recoverable operations and turn database decisions into project evidence for interviews.
Only authorized identities can read protected data.
Controls: access rules, encryption, maskingData changes remain valid, attributable and protected from unauthorized modification.
Controls: constraints, transactions, privileges, auditAuthorized users can access the system when needed.
Controls: redundancy, backups, monitoring, capacityUser or service presents an identity.
Password, key, certificate or federated identity.
Roles and policies permit specific operations.
Relevant activity becomes reviewable evidence.
Grant only required operations, objects and scope.
Avoid giving one identity unnecessary control over development, approval and production.
Unspecified capability remains unavailable.
Remove unused accounts, inherited privileges and temporary access.
CREATE ROLE report_reader;
GRANT SELECT ON monthly_sales TO report_reader;
GRANT report_reader TO analyst_asha;
REVOKE SELECT ON monthly_sales FROM report_reader;Exact syntax, role inheritance, ownership and cascading behavior differ by product. Test revocation: access may still arrive through another role, ownership or a broader privilege.
Avoid using the database owner or administrator account inside an application.sql = "SELECT * FROM users
WHERE email = '" + email + "'";An attacker may supply syntax that changes the intended statement.
' OR '1'='1sql = "SELECT * FROM users
WHERE email = ?";
execute(sql, [email]);The driver sends SQL structure and the value separately, so the value stays data.
For values in SELECT, INSERT, UPDATE and DELETE statements.
Table names, column names and sort directions often cannot be value parameters; map choices to trusted identifiers.
If injection occurs, a restricted application role reduces reachable damage.
Correct escaping is context and product dependent; prepared/parameterized APIs are the primary defense.
Reduces interception between clients, applications and database endpoints.
Protects database files, disks and backups from some forms of offline access.
Useful for highly sensitive fields, with added key-management and query limitations.
Keep credentials out of source code, rotate them and scope each secret to one environment.
Record who performed sensitive actions, what changed, when and from where.
Minimize collected personal data and delete it according to policy and legal obligations.
Avoid recording passwords, access tokens or unnecessary personal values. Protect audit integrity, restrict access and define retention.
Maximum acceptable data loss measured backward from disruption.
RPO = 15 minutesTarget time to restore an acceptable service after disruption.
RTO = 2 hoursComplete backup baseline; simple restore foundation but larger and slower to create.
Captures changes since a defined point; reduces backup volume but adds restore steps.
Uses a base backup plus continuous/archived log records to restore near a chosen moment.
Improves availability but is not a substitute for independent backups; corruption or mistakes may replicate.
Match business impact, not guesswork.
Encrypt, separate access and keep independent/off-site copies.
Verify schema, data, permissions and application behavior.
Record achieved recovery point/time and improve the runbook.
Each project must demonstrate requirements, design, implementation, testing, security and measurable improvement.
Manage students, courses, enrolments, attendance and results.
Coordinate customers, carts, orders, payments and stock.
Protect patients, appointments, prescriptions and clinical notes.
Who needs the system, what workflow fails today and what scope you chose.
ER diagram, assumptions, keys, constraints and relational mapping.
Migration/schema scripts, representative queries, transactions and seed data.
Constraint tests, edge cases, concurrency checks and expected results.
Before/after plan evidence with representative data—not unsupported “faster” claims.
Role matrix, parameterized access, secret handling and audit/backup decisions.
A short reproducible path showing the main workflow and one failure case.
Trade-offs, limitations, what you would improve and what you personally implemented.
“Created a database management system using SQL.”
→EVIDENCE-BASED CLAIM“Designed a 12-table 3NF order schema, protected stock updates with transactions, and reduced a representative report plan from a full scan to a selective composite-index path.”
Select an identity, object and operation. The simulator explains whether access should be allowed and shows the smallest relevant authorization statement.
Educational policy example. Real authorization must also consider row-level rules, ownership, inherited roles and the selected database product.
Choose a blueprint and mark only evidence you can actually demonstrate. Progress is saved in this browser.
ER model → keys → mapping → normalization
DDL → DML → joins → subqueries → windows
Indexes → plans → transactions → recovery
Privileges → injection defense → backup → audit
Mark Level 18 when you can demonstrate security decisions and present one complete database project.
Saved in this browser only.