Disaster Recovery RPO Calculator

Derive the RPO your replication mode and backup interval actually deliver, the data at risk in the worst case, and the interval a target demands.

Inputs

Data protection mode

Backup intervalhours

Typical asynchronous replication lag at peakseconds

Log shipping intervalminutes

Data change rateGB/hour

Time to notice the loss and stop writingminutes

Target RPOminutes

Restores verified from these copies in the last quarter

Achievable RPO

10.8minutes

Achievable RPO in Hours

0.179hours

Data at Risk in the Worst Case

2.15GB

Gap Against the Target

-49.3minutes

Interval the Target Demands

50minutes

Target Verdict

Target met — worst-case data loss sits inside the RPO

Protection Mode Note

Asynchronous replication is bounded by the replication lag, and lag grows under write bursts — size the RPO on the p99 lag, never the average

Verification Note

Restores verified this quarter — an RPO is only real once data has been recovered from the same copy chain

Step by step

Values used
Data protection mode = Asynchronous replication; Backup interval = 24 hours; Typical asynchronous replication lag at peak = 45 seconds; Log shipping interval = 15 minutes; Data change rate = 12 GB/hour; Time to notice the loss and stop writing = 10 minutes; Target RPO = 60 minutes; Restores verified from these copies in the last quarter = Yes
Disaster Recovery RPO
Achievable RPO = replication lag or interval + time to notice. Synchronous gives 0, asynchronous gives lag + detection, log shipping gives the shipping interval + detection, and backup-only gives the full backup interval + detection.
Data at risk
Data at risk = change rate × achievable RPO; the interval a target demands = target RPO − detection time.
Achievable RPO
= 10.8 minutes
Achievable RPO in Hours
= 0.179 hours
Data at Risk in the Worst Case
= 2.15 GB
Gap Against the Target
= -49.3 minutes
Interval the Target Demands
= 50 minutes
Target Verdict
= Target met — worst-case data loss sits inside the RPO

How it works

The detection term is what most RPO calculations omit: your copy is only as good as the last point before the damage, so the minutes before anyone notices count against the objective just as much as the replication interval. That single term is why continuous replication does not automatically give a near-zero RPO against logical corruption, which propagates faithfully to the replica. RPO is where policy and physics diverge most often — nightly backups and a one-hour RPO appear together in real documents. This is an estimate from your own intervals and change rate; only a verified restore proves the recovery point.

Formulas

Disaster Recovery RPO

lagMinutes: Peak asynchronous replication lag converted to minutes
detect: Time before anyone notices and stops the damage spreading
backup interval: Worst case for backup-only protection is a whole interval of changes

Data at risk

Data at risk = change rate × achievable RPO; the interval a target demands = target RPO − detection time.

dataLossGb: Volume of change that would have to be re-entered or accepted as lost
requiredIntervalMinutes: Never larger than the target minus the time to notice

Frequently Asked Questions

How is Disaster Recovery RPO calculated?

Achievable RPO = replication lag or interval + time to notice. Synchronous gives 0, asynchronous gives lag + detection, log shipping gives the shipping interval + detection, and backup-only gives the full backup interval + detection. The detection term is what most RPO calculations omit: your copy is only as good as the last point before the damage, so the minutes before anyone notices count against the objective just as much as the replication interval. That single term is why continuous replication does not automatically give a near-zero RPO against logical corruption, which propagates faithfully to the replica.

Why does Disaster Recovery RPO matter?

RPO is where policy and physics diverge most often — nightly backups and a one-hour RPO appear together in real documents. This is an estimate from your own intervals and change rate; only a verified restore proves the recovery point.

What values do I need to enter?

This calculator takes 8 inputs: Data protection mode, Backup interval, Typical asynchronous replication lag at peak, Log shipping interval, Data change rate, Time to notice the loss and stop writing, Target RPO, Restores verified from these copies in the last quarter. The pre-filled defaults are a realistic starting point — replace them with figures from your own environment for a result you can act on.

Does synchronous replication give me a zero RPO?

Against infrastructure failure, effectively yes. Against ransomware or a bad deployment, no — the replica receives the corruption as faithfully as the good data, so the useful recovery point is your last clean copy. That is why an immutable, independent backup remains necessary alongside replication.

Should I use average or peak replication lag?

Peak, or better the p99. Lag matters precisely when write volume is highest, which is often when incidents happen, and a system averaging two seconds of lag can sit minutes behind during a batch window. Sizing on the average is the most common way an RPO is quietly missed.

You might also need

Disaster Recovery RTO Calculator Business Continuity Calculator Cyber Resilience Calculator