Business Continuity & Business Impact Analysis

Liên tục Kinh doanh & Phân tích Tác động

BCP vs DRP

BCP — Business Continuity Plan

Kế hoạch Liên tục Kinh doanh · PARENT PLAN

Keep business operations running during a disruption. Business-focused — may include manual workarounds, staff redeployment, alternate sites.

Scope: Entire business; includes IT and non-IT functions

DRP — Disaster Recovery Plan

Kế hoạch Phục hồi Thảm họa · SUBSET OF BCP

Recover IT systems and data after a disaster. Technically focused — covers backup restoration, failover, system recovery procedures.

Scope: IT systems and data only; supports BCP execution

Relationship: BCP is the parent plan. DRP is a subset of BCP focused on IT recovery. An organization cannot execute a BCP without the DRP component for IT-dependent business functions.

Business Impact Analysis (BIA) — Key Metrics

The BIA is the foundation of BCP. It identifies critical business functions and defines the recovery parameters for each. These metrics are set by the business — not IT.

Metric	Tiếng Việt	Definition	Who Sets It
MTD Maximum Tolerable Downtime	Thời gian ngừng hoạt động tối đa	Maximum time the business can survive without the function. If exceeded, permanent damage occurs.	Business (executive)
RTO Recovery Time Objective	Mục tiêu thời gian phục hồi	Target time to restore the system/function after an incident. Must be ≤ MTD.	Business + IT jointly
RPO Recovery Point Objective	Mục tiêu điểm phục hồi	Maximum acceptable data loss expressed in time. "How old can the restored data be?"	Business (executive)
WRT Work Recovery Time	Thời gian phục hồi công việc	Time to validate and restore data integrity after systems are back online.	IT (technical assessment)

Critical Relationship — Must Memorize

RTO + WRT ≤ MTD

If RTO + WRT exceeds MTD, the business will fail before IT can restore operations.

RPO Example: If RPO = 1 hour, backups must be taken at least every 1 hour. If a failure occurs, the most data that can be lost is 1 hour of transactions. For a fintech lending platform, an RPO of 15 minutes means backups every 15 minutes or continuous replication.

Recovery Site Types

Site Type	Tiếng Việt	Readiness	Cost	Best For
Hot Site	Trung tâm dự phòng nóng	Immediate failover — fully operational, data mirrored in real-time	Highest	RTO < 1 hour; mission-critical systems
Warm Site	Trung tâm dự phòng ấm	Hours — hardware pre-provisioned, data restored from recent backup	Medium	RTO 1–12 hours; important business systems
Cold Site	Trung tâm dự phòng lạnh	Days — empty facility with power/connectivity; everything must be installed	Lowest	RTO > 24 hours; low-priority systems
Cloud/Elastic Site	Dự phòng đám mây	Minutes to hours — infrastructure provisioned on demand via IaC	Pay-per-use	Variable RTO; modern cloud-native systems

BCP Testing Types (Least → Most Disruptive)

Checklist / Document Review

Review the plan documents to verify completeness and currency. No systems affected.

Structured Walkthrough / Tabletop Exercise

Team discusses scenarios and responses verbally. No systems affected. Cheapest active test. Most commonly used for initial DR validation.

Simulation Test

Simulate a disaster scenario; teams respond as if real, but production systems not actually interrupted.

Parallel Test

Recovery systems activated and tested simultaneously while production continues running normally. Validates failover without interrupting production.

Full Interruption Test

Actual production systems shut down; full failover to recovery site. Most realistic — proves the plan actually works. High risk if plan has gaps.

Key Terms

Term	Tiếng Việt	Definition
BCP	Kế hoạch liên tục kinh doanh	Parent plan for maintaining business operations during disruption
DRP	Kế hoạch phục hồi thảm họa	Subset of BCP focused on IT system recovery
BIA	Phân tích tác động kinh doanh	Foundation of BCP; identifies critical functions and recovery parameters
MTD	Thời gian ngừng tối đa	Maximum time before permanent business damage; set by business
RTO	Mục tiêu thời gian phục hồi	Target restoration time; must be ≤ MTD
RPO	Mục tiêu điểm phục hồi	Maximum acceptable data loss in time (drives backup frequency)
WRT	Thời gian phục hồi công việc	Time to verify data integrity after systems restored
Hot Site	Dự phòng nóng	Immediate failover; most expensive; fully mirrored
Tabletop Exercise	Bài tập bàn	Discussion-based DR test; cheapest; no systems affected

Exam Tips

1. RTO must be ≤ MTD: If RTO exceeds MTD, the business will fail before IT can restore operations. This relationship is critical and frequently tested.
2. BCP = business (keeps running); DRP = IT recovery. BCP is the parent plan. "Which plan governs overall business operations during a disaster?" → BCP, not DRP.
3. Tabletop = cheapest and least disruptive DR test. Full interruption = most realistic and most disruptive (highest risk if gaps exist).
4. Hot site = immediate failover (most expensive); Cold site = empty building (slowest to activate, cheapest).
5. RPO measures data loss in time, not in records. RPO = 1 hour means the system can tolerate losing at most 1 hour of transactions. This drives your backup frequency (backups must be taken at least as often as the RPO interval).

Work Application — FinTech Company X Platform C BCP

Platform C BCP parameters by product:

Product/Partner	MTD	RTO	RPO	Site Type
Partner A VN (live loans)	8 hrs	4 hrs	15 min	Warm (GCP multi-region)
Bank A VN (Platform B cards)	4 hrs	2 hrs	5 min	Warm + active-active
Partner E PH (card, planned)	2 hrs	1 hr	5 min	Hot standby required
Partner D PH (live loans)	12 hrs	6 hrs	30 min	Warm (GCP)

Recommended action: Conduct a tabletop exercise with the Platform A/Platform C engineering team to walk through three scenarios: (1) Kafka broker failure — verify consumer lag alerts fire within 5 min; (2) PostgreSQL primary failover — verify replica promotion RTO; (3) eKYC Vendor eKYC outage — verify manual fallback KYC procedure is documented and team knows it.

Practice Questions

Q1. A bank's loan disbursement system has an MTD of 6 hours. The IT team's RTO is 4 hours and the WRT is 3 hours. What is the problem with this plan?

A) RTO + WRT (4 + 3 = 7 hours) exceeds MTD (6 hours) — the business will fail before recovery is complete

Rationale: RTO + WRT must be ≤ MTD. In this case, 4 + 3 = 7 hours, which exceeds the MTD of 6 hours. The IT team must either reduce RTO (faster restoration), reduce WRT (faster data validation), or negotiate with the business to extend the MTD. This is the most important BCP metric relationship on the exam.

Q2. The Platform C team meets to discuss how they would respond if the primary GCP region went down, talking through their runbook without actually activating any systems. What type of DR test is this?

A) Tabletop exercise (structured walkthrough) — discussion-based test with no systems affected

Rationale: A tabletop exercise involves key stakeholders talking through their response to a disaster scenario without actually activating systems. It is the cheapest and least disruptive form of DR testing, making it ideal for initial validation and training. The next step would be a simulation test (practice without production impact) or parallel test.

Q3. Partner E's planned card processing requires a 1-hour RTO. Which recovery site type is most appropriate?

A) Hot site — immediate failover with data already mirrored; only hot sites reliably achieve sub-1-hour RTO

Rationale: Hot sites have systems fully operational and data mirrored in real-time, enabling immediate failover measured in minutes, not hours. Warm sites require 1–12 hours for data restoration. Cold sites require days. For a 1-hour RTO on a card processing system, only a hot site (or cloud active-active) reliably meets this target.

Q4. A disaster destroys Partner A's loan servicing systems. The CISO activates both the BCP and DRP. Which plan covers resuming loan officer manual activities while IT is restored?

A) BCP — covers all business operations including non-IT processes like manual workarounds

Rationale: The BCP is the parent plan covering all business operations during disruption — including non-IT activities like manual processing, staff redeployment, and customer communication. The DRP is a subset focused specifically on IT system recovery. Manual loan officer activities are a business continuity measure, not an IT disaster recovery measure.

Q5. Platform C has an RPO of 30 minutes for its PostgreSQL lending database. What does this mean for backup frequency?

A) Backups must be taken at least every 30 minutes — the maximum acceptable data loss is 30 minutes of transactions

Rationale: RPO defines the maximum acceptable data loss in time. An RPO of 30 minutes means the organization can tolerate losing at most 30 minutes of data. To guarantee this, backups must occur at least as frequently as the RPO interval. For Platform C's lending database, this likely means continuous WAL archiving (PostgreSQL streaming replication) rather than periodic snapshots.

← 06 Risk Treatment 08 Personnel Security →