You think your cloud backup has you covered. It doesn't. Not by default. The marketing pages say "secure," "reliable," "automatic," but they never mention the recovery point objective you're actually getting, or that your snapshots might be worthless when ransomware hits. I've been in this game long enough to know: cloud backup is only as good as the restore you can actually perform. So let's cut through the fluff and compare three real-world cloud backup strategies head-to-head, because the differences will bite you when it matters most.
The Myth of "Set and Forget" Cloud Backup
The misconception is that once you sign up for a cloud backup service, your data is safe. Wrong. Cloud backup is a tool, not a guarantee. The 3-2-1 rule—3 copies, 2 media types, 1 off-site—is the gold standard, and it's easy to think the cloud satisfies the "off-site" copy (CISA). But the cloud can also be your only copy if you're not careful. And if that cloud provider's snapshot interval doesn't match your recovery point objective, you could lose more than you're willing to accept. NIST SP 800-209 makes it clear: point-in-time copies like snapshots must be configured to meet the RPO of your data. If your RPO is five minutes, a daily snapshot won't cut it.
Contenders: Three Cloud Backup Strategies
- Cloud-to-Cloud Backup: Using a third-party service to back up your SaaS data (Office 365, Google Workspace) to another cloud.
- Cloud Storage Gateways: A local appliance or software that syncs your on-premises data to a cloud provider like AWS or Azure, often with snapshot capabilities.
- Cloud Backup as a Service (BaaS): A provider like Backblaze or Carbonite that installs an agent and continuously backs up your files to their cloud.
Each has its own profile on the three criteria that matter most: RPO, RTO, and ransomware resilience. Let's dig into each.
Criteria: RPO, RTO, and Ransomware Resilience
RPO (Recovery Point Objective) is how much data you can afford to lose, measured in time. NIST defines it as the point in time to which data must be recovered after an outage. If your RPO is 15 minutes, you need backups at least every 15 minutes. RTO (Recovery Time Objective) is how fast you must be back online. For mission-critical systems, NIST often assigns an RTO under 1 hour and an RPO under 5 minutes. But most cloud backup services default to daily or weekly backups, which means your RPO is likely 24 hours or more. That's a problem.
Ransomware resilience is the newest battlefield. Attackers know backups are the last line of defense, so they target them. NIST SP 800-209 warns that attackers interfere with the backup process itself to poison future copies. If your cloud backup is just a mirror of your live environment, ransomware can encrypt both. That's why the 3-2-1-1-0 rule adds an immutable or air-gapped copy and zero unverified backups (CISA).
Comparison: The Table
| Option | RPO | RTO | Ransomware Resilience |
|---|---|---|---|
| Cloud-to-Cloud | Usually 12-24 hours (snapshot-based) | Hours to days (re-download from cloud) | Low—attacker can access both clouds if credentials are compromised |
| Cloud Storage Gateway | Can be near-zero with continuous sync | Hours to days (full restore from cloud) | Medium—snapshots may be in cloud, but if source is encrypted, snapshots may be unusable |
| BaaS | Variable—could be 24 hours or less, but often daily | Hours to days (depends on data size) | Medium—some offer versioning, but not always immutable |
None of these, by default, give you a 5-minute RPO. None give you an RTO under an hour unless you have a plan for rapid restore. And none are inherently ransomware-proof. The sad truth: most cloud backup services are designed for convenience, not for disaster recovery.
Who Wins? Depends on Your Risk Tolerance
For a small business with a low budget, BaaS is the easiest to set up. It's a step up from having no backup, but you're trusting the provider's default schedule and hoping you never need to restore a large dataset quickly. I'd say that's acceptable for non-critical data. For an enterprise with strict compliance requirements, you need more control. Cloud storage gateways with continuous replication can get you close to your RPO, but you'll need to architect the restore process and test it regularly.
The cloud doesn't automatically make your backups better. It's just another storage location. The 3-2-1 rule still applies—you need copies on different media, and one off-site. If your only backup is in the cloud and your provider has an outage (it happens), you're stuck. NIST SP 800-34 requires that your alternate storage site facilitate recovery in line with your RTO and RPO, and that you identify accessibility problems beforehand. Does your cloud provider guarantee that? Read the SLA.
Quick tip: Whatever you choose, enable versioning or point-in-time recovery in your cloud provider. And never keep your encryption keys with your backups—store them separately (NIST SP 800-209).
What I'd Actually Do
I wouldn't rely on any single cloud backup service. I'd use a hybrid approach: a local NAS for fast recovery and a cloud BaaS for off-site protection, with a clear tiering strategy. For critical data, I'd configure continuous backup (like CDP) to meet a 5-minute RPO, and I'd test restores monthly (NIST SP 800-209). I'd also ensure at least one copy is immutable or air-gapped to survive ransomware (CISA). And I'd never claim I have a backup until I've restored from it. Because if you haven't tested it, it's not a backup.
Sources
- CISA - https://www.cisa.gov/stopransomware
- NIST SP 800-209 - https://nvlpubs.nist.gov/nistpubs/SpecialPublications/NIST.SP.800-209.pdf
- NIST SP 800-34 Rev. 1 - https://nvlpubs.nist.gov/nistpubs/Legacy/SP/nistspecialpublication800-34r1.pdf
Comments (0)
Please sign in to post a comment.
Don't have an account? Create one
No comments yet. Be the first to comment!