Server Setup, Backup & Disaster Recovery

We design, deploy and maintain servers for NSW businesses. The default starting point for most businesses we talk to is a cloud-hosted virtual machine in Australia, rather than a physical server in the storeroom. But some workloads still belong on-premise, and we are comfortable with both.

What we set up

• Azure virtual machines in Australia East or Australia Southeast, for workloads that belong in the cloud.
• On-premise servers for workloads that need to stay on-site: POS, manufacturing control, medical imaging.
• Hybrid setups where some data stays on-premise and some replicates to Azure.

Backup and recovery

We follow the 3-2-1 rule: three copies, two media, one off-site. For most businesses, that means a local backup to a NAS, a cloud backup to an Australian provider, and an image backup of the server itself. We test a restore quarterly and give you a written report.

The three backup questions every NSW business should be able to answer

What is backed up?

We meet businesses that think their server is backed up because a tape drive exists in the cupboard. The tape drive has not been used since 2019. We inventory what is actually being protected: file shares, mailboxes, line-of-business databases, configuration and virtual machines.

When was the last successful restore?

If no one in the business can answer this, the backup is theatre. We restore a test file, a test mailbox and a test virtual machine every quarter and produce a written report. The report tells you how long the restore took, what failed, and what was fixed.

How long can the business survive without the server?

This is the recovery time objective. A medical practice cannot be offline for a day. A retail shop can usually work from paper for an afternoon. We size the backup solution to the answer, not to the lowest monthly quote.

On-prem vs cloud-hosted servers

Most of the NSW businesses we talk to do not need a physical server in the office. Azure virtual machines in Australia East are cheaper to run, easier to back up, and survive an office flood. The exceptions are workloads that need to stay on-premise for regulatory, latency or reliability reasons: POS systems that must work when the internet drops, manufacturing control systems, and medical imaging workstations. We are comfortable running either model, or a hybrid that does both.

Disaster recovery planning

A backup is a copy of data. Disaster recovery is a documented plan for getting the business running again after the data centre, the office or the server room is unavailable. We write the plan, test it once a year, and update it when the business changes. The plan is three pages, not fifty, and it is written for the practice manager who will need to follow it at 6am on a Saturday.