This week was a delivery week for the PodHeitor team: we put a Bacula Enterprise environment into production inside one of the main Brazilian public cloud players, doing native backup of OpenStack virtual machines. What closed the acceptance test wasn’t backup — it was restore working across different availability zones of the same cloud.
Sounds like a technical detail. It isn’t. It’s what decides whether a continuity strategy exists or is just a slide deck.
Why cross-restore matters more than the backup itself
In-place restore is easy. Cross-zone — leaving a failed zone behind and bringing the workload up in another zone, in the same cloud or somewhere else — is where most implementations break. The reasons are familiar: hardware-specific dependencies, proprietary image formats, agents that only talk to the original hypervisor, snapshots that turn into platform hostages.
When the customer already lives inside the Brazilian provider’s OpenStack and Bacula can restore a VM into a different zone from the one that ran the backup, the environment moves from “we have backup” to “we have a DR site”. That’s the difference that matters in an audit, in a ransomware attack, in a regional provider outage.
What this unlocks for on-premises customers
The same plugin serves a much broader scenario: Brazilian companies running Hyper-V, VMware, Proxmox, or Nutanix in their own data centers and still without a DR site — because building a second physical site doesn’t fit the budget.
With Bacula Enterprise’s conversion module + PodHeitor’s OpenStack plugin, those customers can use the Brazilian public cloud as a DR target:
- Backup runs on-prem, on the existing hypervisor.
- Replication ships the copy to the provider’s OpenStack tenant.
- DR restore brings the workload up already in OpenStack format, in any available zone.
The practical effect: a Hyper-V customer in São Paulo can spin its workload back up in a Brasília OpenStack zone if the primary data center goes dark. No second site to buy, no Veeam license per replicated VM, no DR locked into a foreign hyperscaler.
Cost, flexibility, resilience — without lock-in
The win is threefold:
- Cost: Brazilian public cloud charges by GB stored, not per protected VM. For most mid-market customers it lands 40–60% cheaper than running a cold secondary data center or paying per-DR-instance licensing.
- Flexibility: backup today on Hyper-V, restore tomorrow on OpenStack — no agent reinstallation, no manual image conversion. Bacula treats conversion as part of the workflow.
- Resilience: with cross-zone restore, the DR site doesn’t hang on a single provider region. The delivery we closed today validated exactly that scenario in production.
Sovereignty isn’t rhetoric. It’s architecture.
Worth saying again: this is Brazilian cloud, with data under Brazilian jurisdiction, a plugin developed in Brazil, LGPD coverage by construction. The U.S. has the Cloud Act. China has the Cybersecurity Law. Any Brazilian company’s data sitting in a foreign hyperscaler is, legally speaking, outside the reach of national sovereignty.
Backup is the last point at which data is still entirely in the company’s hands. Handing that last point over to a foreign provider — even encrypted — gives up a guarantee that the rest of the stack gave up long ago.
With the Bacula Enterprise + PodHeitor plugins + national cloud stack, the customer closes the loop: primary infrastructure wherever it suits the business, backup and DR in Brazilian cloud, perpetual license, auditable code, local vendor.
Innovation moving out of the deck — now
What we delivered today isn’t a POC or a lab. It’s production, with cross-zone restore signed off by the customer. It’s proof that Brazil has, in 2026, every component needed to assemble a business continuity strategy 100% under national jurisdiction — without losing features, without losing flexibility, at competitive cost and with operational independence.
That’s PodHeitor’s thesis: real sovereign backup, built on open-core technology, native plugins, and partnership with the Brazilian cloud. This delivery is one more piece of that thesis materializing in production.
Want to know whether this scenario fits your environment? Talk to the team.
Disponível em:
Português (Portuguese (Brazil))
English
Español (Spanish)