A disaster recovery site is a rack that can run your organisation when your primary location cannot. It sits in a facility with its own power, cooling and connectivity, independent of your office, and holds the systems you would fail over to — replication targets, backup copies, standby servers, and the network and voice services those depend on.
At the Adelaide CBD facility, a disaster recovery footprint commonly starts at a quarter rack — 9RU with a 0.5kW allocation. Access is 24/7 and included, so you can test properly. Councils, multi-site medical practices and schools are the organisations that most often arrive here first.
Why a second site
The failure modes that take out a primary site take out everything sharing its building, its power and its connectivity.
The other server is in the same room
A second server beside the first protects against that server failing. It does not protect against the room flooding, the switchboard failing, a fire on another floor, or the building being closed.
Backups you cannot reach
A backup appliance in the same building as the data it protects shares that building's fate, as does one that cannot be collected because the site is cordoned off.
Power is the common cause
An office UPS carries minutes. A facility with dedicated transformers, dual UPS and N+1 generators carries an extended mains outage.
Connectivity fails independently
If your office internet is down, infrastructure hosted in that office is unreachable regardless of whether it is running. A facility with multiple carriers terminating in it stays reachable when your office tail does not.
Insurance and obligations
Councils, health providers and organisations handling regulated data are increasingly asked to evidence a continuity arrangement. A named facility and a contractual power SLA are straightforward to point at.
Ransomware changes the question
An offline or separately controlled copy in a different location, on different credentials, is one of the few things that reliably survives an encryption event that reaches everything on the production network.
Connectivity to the DR site
Bandwidth and latency between the sites determine whether your RPO is achievable.
Dedicated strands
Physical fibre between your premises and the data centre, with no contention and no imposed bandwidth ceiling. The right answer for storage-level replication and short RPOs, where consistent low latency matters more than headline speed.
250 Mbps – 100 Gbps
Point-to-point circuits delivered end to end at Layer 2 or Layer 3, SLA-backed and monitored. Guaranteed bandwidth for replication windows without operating the underlying fibre yourself.
Replicating over a shared internet service is possible and is what many organisations start with, but the bandwidth available to you varies with everything else using that connection, and an overnight replication window that fits comfortably in January may not in June. A private path removes that variable. Both options are quoted per requirement.
Adelaide deployments
Town of Gawler
Dark fibre, data centre co-location and DDoS-protected internet supporting the council's Smart City initiative — infrastructure moved out of depot buildings never designed to house it. Delivered via public tender.
Read case study →
Orthopaedics SA
Data centre co-location, 1 Gbps fibre across four sites and redundant NBN, eliminating a single point of failure for a multi-site South Australian medical practice.
Read case study →
Mary MacKillop College
Data centre migration, 10 Gbps fibre, backup internet and a hosted phone system ensuring continuity of learning through extended power outages on campus.
Read case study →Rack sizes for a DR site
Three usual starting points. Pricing is fixed for the contract term — see rack pricing.
Smaller than a quarter rack? A backup target that needs one or two rack units can sit in a shared cabinet — see per-RU colocation. The tiers below are the usual starting points where the DR site has to run something, not just hold a copy.
Backup target
Offsite copy, restored from when needed.
9RU with a 0.5kW allocation, including 100 Mbps.
Warm standby
Replicated, ready to start on failover.
20RU with a 2kW allocation, including 100 Mbps.
Full secondary site
Capable of carrying production load.
Full rack, 3kW. Replication link quoted separately.
Setup fees apply on 12-month agreements and are waived on 36-month agreements. Dark fibre and managed Ethernet replication links are quoted per requirement. Rack pricing →
Frequently asked questions
-
A rack in a data centre, separate from your primary site, holding the infrastructure you would fail over to if your main location became unavailable. It has its own power, cooling and connectivity, independent of your office. At Caznet's Adelaide CBD facility a disaster recovery footprint typically starts at a quarter rack, 9RU with a 0.5kW power allocation.
-
A quarter rack with a 0.5kW allocation is the common starting point for replication and backup targets. A half rack at 2kW suits a warm standby. Both include 100 Mbps of unlimited internet, and pricing is fixed for the contract term.
-
Far enough to survive the event you are planning for, close enough that your staff can attend it. For most Adelaide organisations a CBD facility gives useful separation from a suburban head office — different building, different power supply, different physical risk — while staying within a short drive.
-
Backup is a copy of your data. Disaster recovery is the ability to run your organisation from somewhere else. A backup in a rack is not disaster recovery unless something in that rack can run the workload and you have tested restoring onto it.
-
Yes. Dark fibre or managed point-to-point Ethernet from 250 Mbps to 100 Gbps can run between your premises and the data centre on a private path rather than the public internet. For replication traffic this matters — it removes the bandwidth contention and variable latency that make replication windows unpredictable over a shared internet service.
-
A DR footprint carries the workloads you must run in an outage, not everything you run normally, and commonly runs on consolidated or older hardware. A quarter rack is enough for many Adelaide organisations.
More on disaster recovery
Plan a disaster recovery site
Rack space, power and connectivity for a second site in the Adelaide CBD.