Melbourne's storm season, network faults, and ageing infrastructure mean power outages are a real business risk. Here is how to protect your IT systems and keep operating when the lights go out.
Melbourne gets its share of storms. Between the November–March storm season and the occasional infrastructure fault, power outages are a genuine operational risk for businesses in the CBD, inner suburbs, and outer areas alike. Most business owners think about outages in terms of lights going out and the kettle not working. The IT dimension is often more consequential - and more preventable.
Here is a practical guide to protecting your business’s IT systems and maintaining continuity when power fails.
Why Power Outages Are an IT Risk
An unexpected power loss is not just an inconvenience - it can cause active harm to IT systems:
Data corruption. Servers, NAS devices, and workstations that lose power mid-write can corrupt open files and, in some cases, file system structures. While modern file systems have recovery mechanisms, data loss is a real possibility.
Hardware damage. Sudden power loss followed by restoration can expose equipment to voltage spikes (particularly if the power comes back dirty). Power supplies, hard drives, and network switches can fail as a result.
Loss of in-progress work. Any unsaved work open at the moment of power loss is gone. Depending on autosave settings and the application, some or all of a document may be unrecoverable.
Extended downtime for proper restart. Servers do not always restart cleanly after an abrupt shutdown. Checking file system integrity, verifying application health, and bringing services back online in the right order can take an hour or more for a single server - longer for complex environments.
Communication blackout. If your phones run on a PBX that requires power, or your NBN connection equipment loses power, your phone system may go down with the lights.
Uninterruptible Power Supplies (UPS): The Foundation
A UPS is a battery backup system that provides power to connected equipment during an outage. For IT equipment, it serves two purposes: it keeps equipment running during brief outages (long enough for the power to restore), and it provides enough runtime during longer outages to shut down gracefully.
For servers and network equipment, a rack-mounted or tower UPS is essential. Key specifications to consider:
- VA (volt-amperes) rating - must exceed the total draw of connected equipment. Factor in 20–30% headroom.
- Runtime - how long the UPS can sustain full load. For a graceful shutdown, you typically need 5–15 minutes. For any longer runtime, you need a larger battery bank or a generator.
- Automatic voltage regulation (AVR) - cleans incoming power and protects against sags and surges. Worth paying for.
- Network management card - allows the UPS to communicate with your server so it can initiate an automatic graceful shutdown when battery reaches a threshold level.
For a small business server room or rack, budget $800–2,500 for a quality UPS from APC, Eaton, or CyberPower with the features above.
For workstations, a simple desktop UPS ($150–300) provides enough runtime to save work and shut down properly during an outage. This is not always cost-effective for every desk, but it is worth considering for roles where work-in-progress loss would be particularly costly.
For your internet equipment (NBN NTD, router, firewall, switch), a small UPS keeps your connectivity up during brief outages. Given that most NBN services draw 20–40W, even a modest UPS will keep your internet running for several hours.
Configuring Graceful Automatic Shutdown
A UPS is only as useful as its shutdown configuration. Many businesses install a UPS and leave it in its default state - which means it will run until the battery is exhausted and then let the server crash anyway.
Configure automatic graceful shutdown:
- Install the UPS management software (APC PowerChute, Eaton IPM, or equivalent) on your server.
- Connect the UPS to the server via USB or network management card.
- Configure the shutdown trigger - typically when battery reaches 50% remaining or 5 minutes of runtime remaining, whichever comes first.
- Test the shutdown process during a maintenance window by simulating an outage.
For virtualised environments (VMware, Hyper-V), configure the shutdown agent to power off virtual machines in the correct order before shutting down the host. Bring shared services (Active Directory, DNS, file servers) down last and back up first.
Cloud-First Architecture: Your Best Continuity Tool
If your business has moved core applications to cloud services - Microsoft 365, cloud-based accounting, CRM, line-of-business applications - a power outage at your office may affect much less than you think. Cloud applications keep running regardless of your local power status. Staff with laptops (as opposed to desktops) can continue working, provided they have internet connectivity.
This is one of the concrete, practical arguments for cloud-first IT architecture: reduced dependency on local infrastructure means reduced exposure to local infrastructure failures, including power outages.
If you are still running on-premises servers for core applications, a power outage is a good prompt to evaluate whether those workloads make sense in the cloud.
Mobile Internet as Connectivity Failover
When power goes out at an NBN connection point (node or exchange), your internet may fail even if your own office equipment stays up. Having a mobile internet failover - a 4G/5G router configured to fail over automatically when the primary connection drops - keeps your team connected.
In Melbourne, Telstra 4G/5G coverage is reliable enough across metropolitan areas to serve as a business continuity connection for most organisations. A dedicated failover device costs $300–600 plus data, and automatic failover can be configured in most business-grade firewalls (Fortinet, Cisco Meraki, Ubiquiti) without manual intervention.
Generator Considerations
For businesses where extended operation through a long outage is critical - medical practices, 24/7 operations, data-sensitive environments - a generator is worth considering. Key points:
- Standby generators are sized by kVA and need to be correctly matched to your load. An electrician needs to be involved in selection and installation.
- Automatic transfer switches (ATS) detect mains failure and start the generator without manual intervention. This is essential for unattended operation.
- Fuel storage for diesel generators is regulated; check your local council requirements.
- Generator maintenance is critical - a generator that has not been tested under load is not reliable when you need it.
For most Melbourne SMBs, a generator is not necessary if the combination of UPS and cloud architecture means that a multi-hour outage simply means working from home or a café with mobile internet.
Your Power Outage IT Checklist
- UPS installed for all servers, network equipment, and internet infrastructure
- Automatic graceful shutdown configured and tested
- Core applications on cloud services where possible
- Laptops used for staff who need to continue working during outages
- Mobile internet failover in place for connectivity resilience
- Staff know the procedure when power fails (who to notify, how to work remotely)
- UPS batteries tested annually and replaced on schedule (typically every 3–5 years)
Power outages are not preventable - but their impact on your business is manageable with the right preparation. If you would like help assessing your current resilience or designing a continuity solution for your Melbourne business, contact CX IT Services. A short planning conversation now is far less costly than an extended outage later.
