Explore our new Cybersecurity Terms Glossary, explained by our experts.
Cybersecurity Terms Glossary
Cloud backup and recovery is the practice of keeping restorable, tamper-resistant copies of cloud and SaaS data, and proving regularly that you can actually bring a service back.

What Is Cloud Backup and Recovery?
Cloud backup and recovery means keeping copies of your data and configuration in a form that survives whatever happens to the original, and being able to put a working service back within a time the business has agreed to. The second half of that sentence is where most programmes fail. Backups that exist but have never been restored are a filing exercise, not a recovery capability.
Cloud storage is durable. Amazon S3 is designed for eleven nines of object durability, and Azure and Google Cloud offer similar guarantees with geo-redundant options. None of that protects you from the three things that actually destroy data: someone deleting it, software corrupting it, or ransomware encrypting it. Replication copies the mistake faithfully to every region within seconds.
The 2014 collapse of Code Spaces is still the clearest example. An attacker obtained access to the company’s AWS control panel, and when extortion failed, deleted the production environment along with the backups and machine images stored in the same account. The business shut down within days. The technical lesson has not changed: a backup an administrator can delete from the same console is not a backup.
Modern practice centres on copies that nobody—including you—can change for a defined period. S3 Object Lock in compliance mode, Azure Blob immutable storage with a time-based policy, and Google Cloud Storage bucket lock all provide write-once semantics enforced by the platform rather than by policy. Put those copies in a separate account, subscription, or project with its own identity boundary, so that a compromise of the production control plane does not reach them.
The 3-2-1 rule still holds in cloud form: at least three copies, on two different media or platforms, with one of them off-site or logically isolated. Many organisations extend it to 3-2-1-1-0—one immutable copy and zero errors in the last restore test.
Shared responsibility applies to backup as much as to anything else. Microsoft and Google protect the platform and guard against their own infrastructure failures; the contents of your mailboxes, SharePoint sites, Drive folders, and the configuration of your tenant remain yours. Native retention features—retention policies, litigation hold, the recycle bin—are useful but are not designed as backup, and they will not help after a malicious administrator empties them or after a long-running corruption goes unnoticed past the retention window. The same applies to Salesforce, Jira, GitHub, and the rest of the SaaS estate described in our entry on IaaS, PaaS, and SaaS.
Two numbers anchor the design. The recovery point objective is how much data you can afford to lose, which dictates backup frequency. The recovery time objective is how long the service may stay down, which dictates architecture. A four-hour recovery time objective is not met by a nightly snapshot in cold storage that takes six hours to rehydrate, no matter what the policy document claims.
Write the targets per service rather than per organisation, agree them with the business owner, and record the cost of each one. The honest conversation about whether a fifteen-minute recovery point objective is worth its price is far easier to have before an incident than during one.
NCSC guidance on mitigating ransomware is consistent on this point: offline or otherwise isolated backups, tested restores, and a documented plan are what separate a bad week from a terminal one. For organisations in scope of the NIS Regulations or DORA, demonstrable recovery capability is a regulatory expectation rather than a nice-to-have.
A recovery plan is only as good as the last time you proved it. We rehearse cloud recovery alongside incident scenarios through our emergency action plan services.
Microsoft protects its own infrastructure and offers retention, versioning, and recycle bins, but it does not provide a backup of your tenant in the traditional sense. Items deleted beyond the retention window are gone, and a compromised administrator can shorten retention. If you need point-in-time recovery of mailboxes, sites, or Teams data months later, you need a third-party backup or your own export process.
Immutability means the storage platform itself refuses to modify or delete an object until a defined date has passed, regardless of who asks. S3 Object Lock in compliance mode, Azure immutable blob storage, and Google Cloud bucket lock all work this way. Crucially, a compromised administrator account cannot shorten the retention period, which is what makes immutability effective against ransomware.
Test a sample restore monthly and a full service recovery at least once a year, more often for anything with a short recovery time objective. Record the elapsed time and any manual step that was not in the runbook. Most failed recoveries are not caused by missing data but by missing credentials, undocumented dependencies, or a procedure that assumed a system still available.
Yes, wherever practical. A separate account, subscription, or project with its own identity boundary and no trust relationship to production means that compromising the running environment does not automatically compromise the copies. Combine that separation with immutable storage and restricted deletion permissions, and you remove the single most damaging outcome of a cloud control-plane compromise.
You May Be Interested In These
We use a few cookies that are needed to make this website work, and they are always on. We would also like to set analytics cookies, to see which pages are useful; functional cookies, so that our live chat works; and advertising cookies, to measure our campaigns. We will not set any of those until you say yes, and you can change your mind at any time from the Cookie Settings link in the footer.
Choose the cookies you are happy for us to use. Strictly necessary cookies are always on; everything else is off until you turn it on.
These cookies keep the website working: they hold your security checks, your form submissions and the choice you make here. The site cannot run without them, so they are always on.
These let our live chat open and remember your conversation, and let the newsletter sign-up work. Switch them off and the chat window will not appear.
These tell us which pages are read and how people find us, so that we can improve the site. The reports are about groups of visitors, never about you personally.
These let us measure our advertising and show you our adverts on other websites, such as LinkedIn, Google and Meta. Switch them off and you will still see adverts, but they will be less relevant.