UK’s trusted IT infrastructure partner since 2003
sales@servnetuk.com
0800 987 4111
Servnet
ConfiguratorGet in Touch
Disaster Recovery · Buyer's Framework

Choosing a UK disaster recovery provider: the 9 questions that matter

Servnet Editorial · Backup & DR Practice9 min read

UK Disaster Recovery providers — 11:11 Systems (ex-Sungard), Daisy, IPL, IOmart, Pulsant DRaaS, plus Azure Site Recovery + AWS Elastic Disaster Recovery + Veeam Cloud Connect partners — all promise the same thing: "we'll recover you when the worst happens". The differences only show up during a real DR invocation. These are the 9 questions to ask before signing.

DR options — hot, warm, cold compared
Hot siteWarm siteCold / cloud DRRTO< 15 min< 4 hr< 24 hrRPO~ 0< 15 min< 4 hrMonthly costHighestMidLowestTest difficultyEasiestModerateHardestBest forCriticalImportantStd + dev

The 9 questions every UK DR RFP should ask

  • 1. What's your guaranteed RTO per workload tier? Get the number in writing.
  • 2. What's your guaranteed RPO? Continuous replication, hourly, daily, nightly?
  • 3. How many invocation hours per year are included? Most contracts include 4 / 8 / 16 hours; excess hours are expensive.
  • 4. How often do you test? Quarterly tests are standard. Annual tests are insufficient.
  • 5. Show me a real test report. The format + depth tells you a lot.
  • 6. What's your declaration process + decision rights? Who decides we're invoking?
  • 7. Where is the recovery site? Same UK region? Different UK region? Cross-EU?
  • 8. Resource contention — if 5 customers invoke simultaneously, do you have capacity for all of us?
  • 9. Exit assistance — when we leave you, what data + config do we get back?

DRaaS vs cloud-native DR vs traditional DR

Traditional DR (sync replication to your own DR site + colo): highest cost, lowest RTO/RPO, most operational complexity. Best for genuine mission-critical workloads where seconds matter.

DRaaS (provider hosts DR site + replication infrastructure): mid-cost, mid-RTO/RPO, provider operational complexity. Best for mid-market.

Cloud-native DR (Azure Site Recovery / AWS Elastic DR): lowest cost at the right scale, RPO/RTO competitive, requires cloud-native operational maturity.

Backup-based DR (Veeam Cloud Connect / Rubrik Cloud Vault): cheapest but RTO measured in hours-days not minutes. Acceptable for non-critical workloads.

Pick a DR posture per workload tier
What's the max tolerable downtime?
< 1 hr
Hot DR (synchronous repl)
< 8 hr
Warm DR (async + standby)
< 24 hr
Cloud DR (backup restore)

What Servnet does

Servnet partners with UK DRaaS providers + Veeam Cloud Connect + Azure Site Recovery + AWS Elastic DR. We help customers run DR provider selection vendor-neutrally + structure the contract sensibly.

For most UK mid-market customers, the practical answer is: Veeam Backup & Replication with Cloud Connect to a UK provider for tier-2/3 workloads + Azure Site Recovery for tier-1 cloud-native. See our backup platform buyer's guide for the upstream choice.

Key takeaways
  • Get RTO + RPO + included invocation hours + declaration process in writing.
  • Quarterly DR tests are standard; annual is insufficient.
  • Ask for a real test report — format + depth reveal genuine capability.
  • Cloud-native DR (Azure Site Recovery, AWS Elastic DR) is increasingly the right answer for cloud-mature customers.
  • Backup-based DR (Veeam Cloud Connect) is acceptable for non-critical workloads at significantly lower cost.
Frequently asked

FAQs — Choosing a UK disaster recovery provider

Sizing

What RTO / RPO do I actually need?

Tier 1 mission-critical (trading floor, clinical systems, eCommerce): RTO < 1 hour, RPO < 15 min. Tier 2 important (email, file shares, business apps): RTO 4-8 hours, RPO 4 hours. Tier 3 background (archive, dev/test): RTO 24-48 hours, RPO 24 hours. Don't over-pay for Tier-1 RTO/RPO on Tier-2 workloads.

Testing

How do we run a real DR test?

Quarterly tabletop + annual full-invocation test is the gold standard. Servnet supports both — we run the test alongside your team, document outcomes, identify gaps, retest 60-90 days later. Talk to us about scoping.

Related

Got a question this article didn't answer?

One conversation with an engineer who's done this before. No sales script.

Talk to Servnet →