Backup & Disaster Recovery

Backup is the copy; disaster recovery is the plan for using it. Plenty of businesses have backups running and still lose a week to an outage, because nobody decided in advance which systems come back first, how long each one is allowed to take, or who makes the call. We handle both — the copies themselves, and the recovery path and priority order for each system so the decisions are already made when something fails.

That is set per system based on how much work you can afford to redo. The platform supports backups as often as every 15 minutes, and because only changed data is sent, frequent backups do not mean long backup windows or saturated bandwidth. A file server with constant activity might run every 15 minutes while a stable application server runs nightly — we size it during onboarding rather than applying one schedule to everything.

It depends on the system and the recovery method, which is exactly why we choose the method per system in advance. Restoring individual files takes minutes. A full bare-metal restore of a large server takes longer and depends on data volume. For systems where downtime is expensive, we keep a standby image — a ready-to-boot copy in Hyper-V, VMware, or Azure that is refreshed with every backup — so recovery is a boot instead of a restore. We agree those targets with you up front and design to them.

No to both. Backups go direct to cloud, so there is no local backup appliance to purchase, provision, or replace at end of life, and no backup server to patch. Cloud storage and long-term archiving are included rather than metered as a separate bill that grows every month. If you want faster local restores, we can add a second copy on storage you already own, but that is an option rather than a requirement.

Not in the way most people assume. Microsoft guarantees the availability of the service, not the recoverability of your data from your own mistakes — that split is the shared responsibility model, and it is written into your agreement. Retention policies and the recycle bin expire, and they will not help with a deletion discovered months later or a departed employee whose account was removed. We back up Exchange mailboxes, OneDrive, and SharePoint separately, with retention of up to seven years and restores down to a single message or document.

That is the scenario the design is built around. Backups are stored off-site and isolated from your network by default as immutable copies, so ransomware that encrypts your servers cannot reach into the backup storage and encrypt the copies too — the failure mode that turns an incident into a catastrophe when backups sit on a mapped drive or an attached NAS. Access is restricted by role and protected with mandatory multi-factor authentication, and data is encrypted with AES 256-bit encryption in transit and at rest under your own private key.

Because we test them and show you the results. Automated recovery testing restores and boots protected systems on a schedule and captures a screenshot as proof they came back. A backup job that reports success can still be protecting the wrong volume or producing an image that will not boot, and testing is the only thing that catches it before an incident does. For clients under CMMC, NIST 800-171, or HIPAA, that dated test history is also the evidence your assessor asks for.