Tools
Backup and restore time calculator
Most backup plans are judged by whether the job reports success. This tool estimates how long it would take to get everything back with your data size and your links.
Runs in your browser. Nothing is sent.
Result
Full restore time against your target
48 hours of backups. The shaded band is the most data you could lose.
How the backup and restore time calculator works
Network transfers assume 80% of the nominal line speed is usable. 1 Mbit/s moves 0.1 MB per second at that efficiency. Sizes are decimal: 1 TB = 1,000,000 MB.
Compression and deduplication reduce what crosses the network and what is stored. Data written back during a restore is full size, so local restore time uses the full data size.
Worst case data loss (RPO) equals the time between backups. Repository size = (data + daily change × days kept) ÷ reduction ratio, for a forever incremental scheme with one full copy.
Real restores are often slower than the disk or the line allows. Many small files, a busy repository or a single restore stream can double these times. Only a timed restore test gives a real figure.
Planning estimates. They do not replace a measurement on your own systems.
Terms used here: 3-2-1-1-0 ruleDeduplicationImmutable backupRestore testRPORTOSnapshot
In practice
Worked example: 4 TB of data with 2% daily change, compressed 1.5 to 1, over a 100 Mbit/s upload. The first offsite copy takes about three days, each daily copy about an hour and a half, a full local restore about four and a half hours and a full restore from the offsite copy about fifteen hours.
If an eight hour recovery target matters, the offsite restore fails it. The usual answers are a faster download line for emergencies, a local copy that is also immutable, or restoring the most important systems first and the rest later.
Questions about this tool
What are RPO and RTO?
The recovery point objective is how much recent data you can afford to lose, measured in time. The recovery time objective is how long a system can be down. Your backup schedule sets the first and restore speed sets the second.
Why is a real restore slower than the calculation?
Restores rarely run at full disk or line speed. Many small files, deduplicated storage that has to be rebuilt and a single restore stream all slow it down. Plan with a margin and measure with a test.
How often should we test a restore?
At least twice a year for a full system, and monthly for a sample of files. Write down how long it took; that number is your real RTO.
When you want an engineer to look at it
Backup, recovery and virtualisation
We agree how much data you can afford to lose and how long each system can be down, design backup and virtualisation around those targets, and prove them with a timed restore test. Planned moves from VMware to Proxmox or Hyper-V are included.
Tools
Other tools
Infrastructure check
Ten questions on backup, access, power and support. A score out of ten and your three weakest points.
Downtime cost calculator
What an outage costs in idle staff, lost revenue and recovery work, per hour, per outage and per year.
Uptime and SLA calculator
Turn 99.9% into hours of downtime, then build your own chain of components and find the weakest link.