ISC · Information Systems and Data Management · 8 practice questions
RTO vs RPO: Compare Replication Lag and Recovery Time
RPO limits data loss; RTO limits downtime, and a recovery design must meet both. Below, compare Delta's warm-site baseline with changes to replication lag or tested recovery time.
Try one first
Hint
Separate the RPO (maximum acceptable data loss) from the RTO (maximum acceptable downtime); determine whether each proposed technology's replication/restore behavior can reliably stay within both numeric thresholds.
Answer A. Synchronous replication writes committed transactions to both sites before acknowledging them, effectively eliminating data loss and thus meeting the 15-minute RPO. Combined with automated failover and documented recovery within 30 minutes, this solution also meets the 1-hour RTO.
Why not B: Tempting because automated failover and a sub‑hour RTO are provided, but the average 20-minute replication lag exceeds the 15-minute RPO; therefore it does not reliably meet the data-loss requirement.
Why not C: Tempting because tape backups are familiar and reliable for long-term retention, but nightly backups produce roughly a 24-hour RPO and multi-hour manual restores, so this option fails both the 15-minute RPO and the 1-hour RTO.
Why not D: Tempting because 10-minute snapshots meet the 15-minute RPO, but recovery is manual and typically takes 2-3 hours, exceeding the 1-hour RTO, so this option does not satisfy both requirements.
Side by side
| Baseline | Variation 1: Change only replication lag | Variation 2: Change only recovery time | |
|---|---|---|---|
| Unchanged setup | Delta: separate warm site, asynchronous replication, full-site outage. | Same as baseline. | Same as baseline. |
| Required RTO | 1 hour | 1 hour | 1 hour |
| Required RPO | 15 minutes | 15 minutes | 15 minutes |
| Reliable maximum replication lag | 15 minutes | 20 minutes | 15 minutes |
| Tested end-to-end recovery time | 45 minutes | 45 minutes | 75 minutes |
| RPO test | Pass: 15 minutes equals the limit. | Fail: 20 minutes exceeds the limit. | Pass: 15 minutes equals the limit. |
| RTO test | Pass: 45 minutes is within 1 hour. | Pass: 45 minutes is within 1 hour. | Fail: 75 minutes exceeds 1 hour. |
| Meets both objectives? | Yes. | No. Only RPO fails. | No. Only RTO fails. |
Key points
- Remember: Faster recovery does not fix excessive data loss, and fresher data does not fix slow recovery.
- Use reliable maximum replication lag, not average lag, to evaluate the data-loss limit.
- Each variation returns to the baseline and changes only one fact.
How the exam traps you
- Accepting 20-minute replication lag because recovery takes only 45 minutes. Check RPO separately: 20 minutes exceeds Delta's 15-minute data-loss limit.
- Accepting 75-minute recovery because replication lag stays within 15 minutes. Check RTO separately: 75 minutes exceeds Delta's 1-hour downtime limit.
- Treating asynchronous replication as an automatic RPO failure. Compare its reliable maximum lag with the allowed loss. The baseline's 15-minute lag meets the 15-minute RPO.
Question 2
Hint
Focus on whether the site can process transactions immediately, or whether meaningful setup and restoration are still required.
Answer B. A warm site has hardware, connectivity, and an operating environment already in place but typically requires data restoration and some configuration before it can resume operations. The scenario describes servers and network access on-site but still needing backup restores and setup, which is the hallmark of a warm site. A hot site, by contrast, would be ready for immediate or near-immediate processing.
Why not A: This is tempting because servers and network connectivity exist at the facility, but a hot site is generally ready for immediate processing with data and configuration already in place. Here, additional restoration and configuration are required, so it is not a hot site.
Why not C: A cold site typically provides only space and basic utilities (power, cooling, network access may be minimal) without preinstalled servers or a ready operating environment. Because this facility already has compatible servers and the operating environment, it is more capable than a cold site.
Why not D: A reciprocal site refers to an agreement to use another organization's facilities in a disaster; it does not describe the level of preinstalled infrastructure. The facts describe a dedicated recovery facility with its own installed servers and environment, so 'reciprocal site' is not the correct classification.
Question 3
Hint
Focus on what hot, warm, and cold sites are meant to solve during a disruption, not on accounting or staffing details.
Answer B. Selection among hot, warm, and cold sites is driven by business continuity requirements, primarily the recovery time objective (RTO) and recovery point objective (RPO). Systems that must be restored quickly with minimal data loss typically require a hot or warm site, while systems tolerant of longer outages can use colder sites. Cost and other operational factors are secondary and should be considered after defining the required recovery capability.
Why not A: This is tempting because expensive or older hardware might seem to merit greater protection. However, disaster recovery site type is not determined by asset age or accounting value; it is driven by how quickly operations must resume and how much data loss is acceptable.
Why not C: High user counts may indicate importance, but user count alone does not define recovery requirements. The primary consideration is the business's tolerated downtime and data loss (RTO/RPO), not simply how many people access the system.
Why not D: Budget affects what the company can afford, but it should not be the primary governing factor. The business continuity needs (required recovery time and allowable data loss) should determine the required site capability, and budgetary planning follows from those requirements.
Question 4
Hint
Separate controls that improve backup quality from controls that meet both site-failover speed and transaction-loss limits.
Answer A. The requirement combines a site-wide disaster scenario with both a tight RTO of 1 hour and a tight RPO of 15 minutes. A preprovisioned secondary site with continuous or near-real-time replication is the infrastructure approach that most directly addresses both thresholds. Backup-based recovery or single-site redundancy may help resilience, but they do not best satisfy rapid recovery from total site loss with minimal data loss.
Why not B: This is tempting because it uses a secondary site and sounds like a disaster recovery solution. However, hourly replication allows up to 60 minutes of data loss, which misses the stated 15-minute maximum loss threshold even if the failover location is ready. Candidates often focus on the recovery-site part and underweight the replication-frequency requirement.
Why not C: This is tempting because it improves availability and reduces local hardware single points of failure. But the stem already says the concern is the entire primary site becoming unavailable, and extra redundancy inside that same site does not solve site-level disaster recovery. It addresses high availability within a site, not continuity after total site loss.
Why not D: This is tempting because immutable backups and restore testing are strong controls and are often recommended for ransomware resilience. Still, they are primarily backup integrity and recoverability measures, not the fastest way to meet a 1-hour restart objective with only 15 minutes of allowable data loss after losing an entire site. Nightly backups are especially inconsistent with the required data-loss threshold.
Question 5
Hint
Compare each option's replication frequency and failover automation against the stated RTO/RPO, and factor in the small IT team and desire for managed services.
Answer B. Option B balances the RTO/RPO requirements with operational capacity and cost: near-real-time managed replication plus automated failover can meet a 15-minute RPO and 30-minute RTO, while a hot/warm managed standby reduces operational burden on a small IT team and avoids the full cost and complexity of active-active duplication. Keeping data within U.S. regions satisfies residency needs.
Why not A: This offers very high availability and can meet tight RTO/RPO goals, but active-active is more complex and costly to operate continuously. For a small IT staff and moderate budget, it is likely overkill compared with a managed hot/warm standby.
Why not C: On-premises synchronous replication can technically meet the objectives but requires significant capital, staffing, and operational expertise. With only three IT staff and a preference for managed services, this option is less practical.
Why not D: Nightly backups and manual failover cannot meet a 15-minute RPO or a 30-minute RTO; this low-cost approach would expose the company to unacceptable data loss and downtime.
Question 6
Hint
Separate recovery time from recovery point. Ask what automation improves, and what asynchronous replication still allows to happen.
Answer C. Because replication is asynchronous, the DR site can lag behind the primary site even if failover is automated and the DR application servers are ready. If the primary site is lost 4 minutes after the last successful replication point, transactions committed during that 4-minute gap may not yet exist at DR. Automated failover mainly reduces recovery time, not the replication gap (recovery point).
Why not A: This mixes correct-sounding storage terms with an incorrect conclusion. Asynchronous replication can preserve a consistent copy but does not guarantee that the most recent commits reached DR before the outage; also the outage length depends on the failover process (2 minutes), not the replication interval.
Why not B: This confuses compute readiness with data currency. Pre-provisioned application servers reduce recovery time but do not eliminate asynchronous replication lag on the database, so some recently committed transactions can still be missing.
Why not D: That describes synchronous-replication behavior. In asynchronous replication, the primary can acknowledge commits before those changes arrive at the DR site, so some committed transactions may be absent after failover.
Question 7
Hint
Compare the RPO/RTO targets to the replication mode (synchronous vs asynchronous) and whether the design tolerates a full site loss without long manual recovery steps.
Answer B. An active-active design with synchronous replication ensures that committed transactions are present at both sites (near-zero RPO). Automatic failover with a global load balancer can redirect traffic immediately, enabling the RTO to meet the ≤1-minute target. The provided low-latency fiber link makes synchronous replication practical without unacceptable transaction latency. Other architectures either risk recent-commit data loss (asynchronous replication) or require much longer recovery steps.
Why not A: This is tempting because it lowers cost, but asynchronous replication can leave very recent committed transactions only on the primary (violating the near-zero RPO). A cold standby and manual/semi-automated failover typically produce an RTO well above 1 minute.
Why not C: RAID and backups protect against local hardware failure and corruption but do not provide site-level redundancy for a full data center loss; restoring from backups or snapshots will far exceed a 1-minute RTO and can result in data loss for recent transactions.
Why not D: Cloud multi-region deployments can be resilient, but cross-region replication is often asynchronous and DNS-based failover can be slower than 1 minute because of propagation/TTL and routing changes; unless the cloud configuration explicitly uses synchronous replication and an immediate failover mechanism, it may not meet the stated RPO/RTO.
Question 8
Hint
Test each choice against both thresholds separately: one for downtime and one for allowable data loss after a full site outage.
Answer B. The requirement combines a very short recovery time objective (30 minutes) and an almost-zero recovery point objective (no more than 1 minute). Synchronous replication ensures transactions are committed to the secondary site concurrently, minimizing committed-data loss to essentially zero. Preconfigured failover capacity and tested failover procedures support rapid switchover to meet the 30-minute RTO. The other options either allow more than one minute of data loss or lack the preconfigured capacity and tested processes needed for such a fast restore.
Why not A: Nightly backups provide offsite protection but create an RPO measured in hours, far exceeding the 1-minute limit, and rebuilding systems from backups after a full-site loss is unlikely to restore Tier 1 services within 30 minutes.
Why not C: Internal redundancy and daily backups protect against localized failures but do not address a complete data-center loss; daily backups yield an RPO of up to 24 hours and there is no preconfigured secondary capacity or tested failover to meet a 30-minute RTO.
Why not D: This is the closest distractor because it addresses offsite recovery and a warm site, but asynchronous replication with 5-minute intervals permits up to five minutes of committed-transaction loss, which violates the stated 1-minute RPO even if the warm site could meet the RTO.
Common questions
What is the difference between RTO and RPO?
RTO is the maximum acceptable downtime. RPO is the maximum acceptable data loss measured in time; Delta allows 1 hour of downtime and 15 minutes of data loss.
Can asynchronous replication meet a 15-minute RPO?
Yes. Delta's reliable maximum lag of 15 minutes meets its RPO, but increasing only that lag to 20 minutes causes an RPO failure.
Can a warm site meet a 1-hour RTO?
Yes, if tested end-to-end recovery finishes within 1 hour. Delta's 45-minute baseline passes; changing only recovery time to 75 minutes causes an RTO failure.
Practice ISC like the real exam
The free ChatCPA simulator: real exam layout, timed testlets, starting with a question on this topic. No account needed to start.
Open the free simulator →More on IT infrastructure
Questions from the ChatCPA bank of 17,658 CPA exam questions, each with a written reason for every wrong answer. ChatCPA is built by Nicholas Miller, CPA (Oregon #14907). How these pages are made. Spot an error? Tell us.