The questions, including the awkward ones.

Operations. Datastore sizing, PBS and OS patching, disk and SMART health, prune schedules, garbage-collection windows that do not collide with your backup window, verify jobs and the alerting when they fail, offsite sync, backup of the PBS configuration itself, certificate expiry, and a restore test somebody actually runs. The software is free; none of that is.

Offsite PBS if you keep your PVE and want a genuine second copy somewhere else. On-prem managed PBS if the data cannot leave your building and you want us to operate it in place. DR on Proxmox is a separate product with a warm standby cluster and a tested failover — it is deliberately not bundled into the backup price, because bundling it is how backup vendors lose money and customers get promises nobody costed.

Yes. The discipline is identical; only where the disks live changes. On-prem tends to suit BFSI, healthcare, public sector, and anyone with a hard residency constraint.

It earns its keep from roughly twenty protected VMs upwards, and fastest where nobody on staff owns storage or backup full time. Below that, self-run PBS plus a disciplined calendar reminder is usually the honest recommendation, and we will give it.

No, and we previously said otherwise on this site, which was wrong. PBS has no immutability flag, no retention-lock API, and no WORM certification. Namespaces are an organisational boundary, not a lock. What genuinely resists ransomware is the content-addressed chunk store that a backup run cannot overwrite, ZFS snapshots on the datastore, backup tokens issued without prune or delete rights, and optionally S3 Object Lock on a synced bucket.

You do. PBS performs client-side encryption before chunks leave your network, so we hold ciphertext and never the key. The consequence is real and worth stating: lose the key and nobody, us included, can recover those backups. Key custody is a duty we help you plan rather than quietly assume.

PBS listens on TCP 8007. Your PVE hosts connect outbound to it. Nothing needs to be opened inbound towards your cluster.

It can, which is why jobs are rate-limited where the uplink is shared with production. The first run is a full copy; after that PBS sends only changed, deduplicated chunks.

No. We use a PBS API token scoped to your datastore, and depending on the change-control model you choose it does not carry prune or delete rights — so a compromised PVE host cannot erase its own backup history.

Yes. It is a normal PBS datastore attached to your cluster. Full VM, container, and file-level restores work the way they always do. A service that makes you raise a ticket to get your own data back is not one you want on the worst day of your year.

We are, by roughly two to five times on headline per-TB rates, and we would rather you heard that from us. Providers at EUR 8 to 21 per TB are selling storage with a PBS interface in front of it. Our fee covers the operations, run by Proxmox Silver Partner engineers in your timezone with copies held in India. If what you need is cheap bytes and you are happy to run GC, verify, prune, and restore drills yourself, buy the cheap bytes. That is a legitimate choice.

Because protected capacity, retention, whether the PBS is hosted or on-prem, and whether DR is in scope change the number materially. The published Starter and Professional tiers give you the shape; anything larger gets a fixed quote after the health check rather than a range that moves later.

Ask for the exit terms in writing before you sign — from us or from any backup vendor. The arrangement we recommend is an export window during which backups are handed over or synced to a PBS instance you own. “Cancel anytime” without a documented exit path is a warning sign rather than a feature.

Service levels are defined in contract for managed plans. We are deliberately not printing numbers on a marketing page that operations have not signed up to deliver — a published figure nobody is accountable for is worse than no figure at all. Ask during the health check and you will get the specific commitments for your scope.

Computer Port's customer list is on the home page and spans manufacturing, BFSI, public sector, and research. A published restore-test case study specific to this managed backup service is something we are working on, and we would rather say that than dress a general infrastructure reference up as a backup one.