RPO and RTO
RPO is how much data you can lose; RTO is how long you can be down
The recovery point objective (RPO) sets backup frequency; the recovery time objective (RTO) sets restore speed; both sit inside the MTD.
Two numbers drive every recovery plan. Here is a day. A backup runs at two in the morning. At two in the afternoon, the database fails. Everything since the backup is gone: twelve hours of orders. That is your real recovery point. But the business agreed it could lose at most one hour. So backups must run at least hourly. That is the recovery point objective. Service comes back at six: four hours down. That is the recovery time. RPO looks back from the failure. RTO looks forward.
Backup types
Incremental is quick to take and slow to restore; differential the reverse
Full, incremental and differential backups trade nightly size against restore effort; snapshots add frequent copies.
Three kinds of backup, one week. Sunday is always a full backup: five hundred gigabytes. Incremental backups copy only what changed since the last backup: twenty gigabytes a night. Quick. But restoring on Saturday needs the full plus all six incrementals, and if one is bad, the chain breaks. Differential backups copy everything since the last full, so they grow every night, to a hundred and twenty gigabytes. But a Saturday restore needs just two sets. Snapshots and replication add copies every few minutes.
Backup and replication methods · Going deeper
Snapshots, replication, journaling and protecting the backups
Snapshots, replication and journaling shrink the RPO; backups need encryption, separate credentials and test restores.
Snapshots, replication and journaling shrink the RPO; backups need encryption, separate credentials and test restores. Snapshot a point-in-time image of a disk or virtual machine, taken in seconds. Usually stored on the same system, so it is not a backup on its own. Replication continuous copying to another site. Excellent RPO, but it also copies corruption and deletions, so it complements backups rather than replacing them. Journaling records every change so a database can be rolled to any moment, such as one minute before an attack. On-site and off-site on-site restores fast; off-site survives the loss of the building. Most designs use both. Protecting the backups.
- Snapshot a point-in-time image of a disk or virtual machine, taken in seconds. Usually stored on the same system, so it is not a backup on its own.
- Replication continuous copying to another site. Excellent RPO, but it also copies corruption and deletions, so it complements backups rather than replacing them.
- Journaling records every change so a database can be rolled to any moment, such as one minute before an attack.
- On-site and off-site on-site restores fast; off-site survives the loss of the building. Most designs use both.
Protecting the backups
The 3-2-1 rule
Keep one copy that attackers can’t reach, and prove it restores
The 3-2-1-1-0 rule: three copies, two media, one off-site, one offline or immutable, zero restore errors.
How many copies is enough? The classic rule: three copies of the data, on two different kinds of media, with one off-site. Ransomware added a fourth number: one copy offline or immutable, which nobody can change or delete, even with stolen admin rights. In 2024, ninety-four percent of ransomware victims said the attackers went after their backups. And a fifth: zero errors when you test a restore. A backup that has never been restored is a hope, not a plan.
Availability and redundancy
Each extra nine cuts downtime tenfold and costs far more
Availability targets (99.9%, 99.99%) need redundancy: RAID, clusters, load balancers and multiple zones. RAID is not a backup.
Availability is measured in nines. Ninety-nine percent sounds high, but it allows more than three and a half days of downtime a year. Three nines: under nine hours. Four nines: about fifty-three minutes. Five nines: about five minutes, and multiple regions to get there. Each nine comes from removing single points of failure: mirrored disks, dual power supplies, clusters, load balancers, more zones. And remember: RAID protects against a failed disk, not a deletion or ransomware.
Recovery sites
The faster you must recover, the more you keep running in advance
Cold, warm and hot sites, and cloud standby, trade cost against recovery time; keep geographic dispersion.
Where do you recover if the data centre is gone? A cold site: space, power and network. You bring everything else, so it takes weeks, cheaply. A warm site: equipment and somewhat older data, back in hours to days. A hot site: a live duplicate, back in minutes, at the highest cost. And a cloud region: standby capacity you pay for mainly when you fail over. Whichever you choose, put it far enough away that one disaster cannot take out both.
Business impact analysis
The business, not IT, decides what comes back first
A business impact analysis sets MTD, RTO and RPO per process and drives the BCP and the DRP.
Who decides that payments come back before the wiki? The business, through a business impact analysis. Each process owner says how long they can survive without it: the maximum tolerable downtime. Card payments, four hours. The website, a day. Payroll, three days. The wiki, two weeks. Recovery time and recovery point are set inside those limits, and the order follows. The business continuity plan keeps the company working, with people and workarounds. The disaster recovery plan restores the technology.
Continuity planning · Going deeper
BCP, DRP, COOP and the metrics around them
Know BCP, DRP, MTBF, MTTR, succession planning and failover.
Know BCP, DRP, MTBF, MTTR, succession planning and failover. Business continuity plan (BCP) keeps essential functions going during a disruption: alternate locations, manual workarounds, communication, people. Disaster recovery plan (DRP) restores IT systems and data after a disaster, in the order the BIA sets. Continuity of operations (COOP) the government term for keeping essential functions running. Succession planning who decides if key people are unavailable. Reliability metrics. Power counts too: an uninterruptible power supply (UPS) bridges seconds to minutes; generators run for days with fuel contracts in place.
- Business continuity plan (BCP) keeps essential functions going during a disruption: alternate locations, manual workarounds, communication, people.
- Disaster recovery plan (DRP) restores IT systems and data after a disaster, in the order the BIA sets.
- Continuity of operations (COOP) the government term for keeping essential functions running.
- Succession planning who decides if key people are unavailable.
Reliability metrics
Power counts too: an uninterruptible power supply (UPS) bridges seconds to minutes; generators run for days with fuel contracts in place.
Testing recovery
A recovery plan nobody has tested is only a hope
Test with tabletop exercises, walkthroughs, simulations, parallel tests and full interruption tests.
Plans fail in ways nobody predicted, so test them. A tabletop exercise: the team talks through a scenario around a table. Cheap, and it finds gaps in roles and decisions. A walkthrough: each team steps through its own part. A simulation acts out a realistic scenario without touching production. A parallel test brings up the recovery site alongside production, proving the technology. And a full interruption test actually fails over. The most realistic, and the riskiest.
Lesson complete
You finished “Backup and recovery”
Nicely done.