High Availability VPS Australia: What True Uptime Actually Takes
22 Jun 2026, by James Braunegg, CEO and Founder, Micron21
Everything we have built at Micron21 over the past two decades has been designed around one idea: uptime. Our Tier IV certified data centre is designed for uptime. Our global network, AS38880, is designed for uptime. Our DDoS protection is designed to keep you online when someone is actively trying to take you down. But where all of that engineering truly comes together for the customer is our cloud platform, mCloud, and specifically the high availability VPS and cloud server services it delivers.
If you are searching for a high availability VPS in Australia, you have probably already learned the hard way that the words "VPS" and "cloud server" get used very loosely. Almost everyone claims high availability. Very few platforms are actually built for it. This article explains what real high availability looks like, why it is so much harder than it sounds, and how we deliver it on mCloud, all the way up to a genuine triple site, real time replicated cloud server that keeps the same public IP address no matter which data centre it is running in.
What high availability actually means
High availability, usually shortened to HA, is the ability of a system to keep running with minimal interruption even when individual components fail. It is measured in uptime, expressed as a percentage, and the gap between those percentages is far more dramatic than it first appears.
| Uptime | Name | Downtime per year | Typical use |
|---|---|---|---|
| 99.9% | Three nines | About 8 hours 45 minutes | General business applications |
| 99.99% | Four nines | About 52 minutes | Payments, auth, core infrastructure |
| 99.999% | Five nines | About 5 minutes 15 seconds | Life critical, heavily regulated systems |
The important lesson is that those extra minutes are not won by buying a faster server. They are won by removing single points of failure, so that when something breaks, and hardware always eventually breaks, your workload simply keeps running somewhere else. That is the difference between a cheap VPS and a true high availability cloud server.
Why most VPS platforms are not highly available
A virtual private server can mean almost anything. At the budget end, a VPS is a virtual instance running on a single physical host with no redundancy at all. The provider packs as many customers as possible onto one machine, and if that machine fails, every customer on it is offline until the hardware is physically repaired or replaced. There is no automatic failover, because there is nowhere to fail over to. You only discover this on the day it happens, which is always the worst possible day.
True high availability requires the opposite design philosophy. Compute has to be spread across multiple physical servers, so the loss of any one host just means your instance restarts on another. Storage has to be distributed across multiple devices, so the loss of a disk or even a whole storage node never loses your data. The network has to be able to move your IP address to wherever your instance is now running, instantly and without you reconfiguring anything. Get all three right and you have a platform that absorbs failure quietly. Miss any one of them and you have a single point of failure waiting to take you offline.
mCloud: a true fault tolerant cloud server platform
mCloud is our answer to that problem, and it is built as a true fault tolerant platform rather than a marketing label bolted onto cheap hosting. It comes in two levels of resilience, so you can match the investment to how much downtime your business can actually tolerate.
The first is single site redundancy. Here your VPS runs inside our Tier IV data centre on a platform where everything is fully redundant: multiple compute servers, multiple storage devices, redundant networking, redundant power and cooling. If any single component fails, your instance keeps running and the platform self heals around the fault. For the overwhelming majority of Australian businesses, this delivers the high availability they need, inside the most resilient data centre classification there is.
The second is for organisations that cannot tolerate even a site level event, and this is where mCloud does something most Australian providers simply cannot. With triple site redundancy, we replicate your virtual instance in real time across three separate Melbourne data centres: our own Tier IV facility, NextDC, and Vocus, with your backups stored independently in Equinix. If an entire data centre were to go dark, your VPS fails over to another site almost instantly, with your data already there waiting. This is not a nightly backup that you restore after an outage. It is live, multi site replication designed so that the loss of a whole building is something your customers never even notice.
What ties this together, and the part we are most proud of, is the network design. Regardless of which of the three sites your instance is running in, the same public IP address remains reachable. We achieve this with a Cisco leaf and spine fabric running EVPN with BGP, using Cisco Nexus anycast gateway services. In plain terms, the same gateway and the same IP live in all three locations at once, so when your workload moves between data centres there is no DNS change, no IP change, and no manual intervention. The network resilience is built into the fabric itself rather than depending on slow failover scripts. For a high availability VPS in Australia, that seamless IP mobility across sites is the feature that genuinely sets the platform apart.
Built end to end, with no external VMware licensing
mCloud is powered by OpenStack, and it delivers a complete cloud server platform that competes directly with AWS, Azure, Google Cloud and the local Australian providers. The difference is that we own everything end to end. The global network, the DDoS protection, the SOC and the NOC, the Tier IV data centre and the cloud platform itself are all ours. There is no reselling, no hidden third party, and no offshore dependency.
That end to end ownership extends to the software stack, and this is where the VMware question matters. Many VPS and private cloud platforms are built on VMware, which historically meant excellent technology carrying significant per socket and per VM licensing costs. Since Broadcom's acquisition of VMware, those costs and licensing terms have changed dramatically, and a large part of the industry is now actively looking for alternatives. Because we built mCloud from the ground up on OpenStack, there is no external VMware licensing baked into your bill, which is one of the reasons we can deliver enterprise grade high availability at a competitive price. OpenStack gives us the same core capabilities enterprises rely on, live migration for zero downtime maintenance, automatic restart of instances on healthy hosts, distributed storage and full software defined networking, without the licensing tax.
To be clear, this is not an argument that VMware is bad. It is excellent software. For customers who specifically want to run VMware, our virtual data centre and bring your own hardware options let them do exactly that on dedicated nodes within our platform. The point is simply that you are not forced into a particular hypervisor licensing model just to get high availability from us.
Underneath the platform, storage runs on Ceph across four distinct tiers, from high performance NVMe, to SAS, to bulk storage, to encrypted NVMe for sensitive workloads. Ceph is itself a distributed, self healing storage system, which means your data is replicated across multiple devices by design. You choose the tier that fits each workload, balancing performance and cost, while the platform handles the redundancy for you.
More than a VPS: virtual data centres and your own hardware
Not every customer wants a single VPS. Some need an entire environment, and mCloud scales to meet that. With our virtual data centre service, we give you a pool of resources, compute, memory and storage, and you decide how to carve it up across as many virtual instances as you like. You become the administrator of your own slice of the cloud, building and resizing servers on demand within the capacity you have reserved.
For organisations with specific performance, compliance or scale requirements, you can also have dedicated nodes, or bring your own hardware into our cloud platform. This gives you the operational benefits of mCloud, the network, the security, the self service control and the high availability design, while running on compute that is exclusively yours. As your business grows, you can move fluidly between shared cloud servers, dedicated nodes and full virtual data centres without ever leaving the platform. The options are effectively unlimited, and they scale with you.
Total control through a self service portal and full API
High availability is only half the story. The other half is control, and a platform you cannot manage quickly is a platform that slows your business down. mCloud gives you complete API access to every function, alongside a full self service portal where you manage the entire platform yourself.
From the portal you can resize an instance, adding more RAM, more CPU or more storage in seconds, take backups and snapshots, add and configure networks, build firewalls and define firewall rules, add load balancers, deploy a web application firewall, and access the console of any instance directly. Everything that would once have required a support ticket and a wait is now something you do yourself, in seconds, at any hour. Because every function is also exposed through the API, you can automate all of it, building infrastructure as code, scaling in response to demand, or integrating mCloud directly into your own deployment pipelines. For development teams this is the difference between infrastructure that gets in the way and infrastructure that disappears into the background.
A VPS uptime guarantee in Australia, backed by real engineering
A VPS uptime guarantee in Australia is only as good as the infrastructure and the people standing behind it. Put simply, a VPS uptime guarantee Australia wide should be measured by the engineering underneath it, not by the number printed on the contract. An SLA promising four nines means nothing if the platform underneath it has a single host that can take you offline, or if there is no one watching at three in the morning. This is why our uptime commitments rest on the full stack we own: a Tier IV data centre with full power and cooling redundancy, the AS38880 network peering with more than 2,000 networks, our own DDoS mitigation absorbing attacks before they reach you, and our SOC and NOC teams monitoring everything around the clock. As we often say, you cannot operate a Tier IV data centre and not secure the network at the same time, and the same logic applies to uptime. Real high availability is the product of every layer working together, not a number written on a page.
Where high availability earns its keep
The value of a high availability VPS becomes obvious the moment downtime has a real cost. For an online retailer, an hour offline during a sale is lost revenue you never recover, so triple site replication with seamless IP failover means a data centre incident does not become a trading halt. For healthcare and government systems, where availability obligations sit alongside data sovereignty requirements, a fault tolerant Australian cloud server keeps critical services running and keeps the data onshore at the same time. For SaaS companies, your customers judge you on your own uptime, so building on a platform that already delivers four nines lets you make credible promises to them. For financial services and payment platforms, where authentication and transactions cannot stop, the combination of distributed compute, distributed storage and anycast networking is exactly the architecture regulators expect to see. And for any business running line of business applications, ERP, databases or internal tools, high availability quietly removes the outages that would otherwise eat into productivity and trust.
In every one of these cases the common thread is the same. The organisations that invest in real high availability are the ones for whom being offline is not an inconvenience but a genuine business risk, and they want that risk engineered away rather than apologised for after the fact.
Build on a platform designed for uptime
A high availability VPS is not about a single powerful server. It is about an entire platform designed so that no single failure can take you down, backed by a company that owns and operates every layer of that platform inside Australia. That is what mCloud is built to be, from single site redundancy in our Tier IV data centre all the way to real time triple site replication across Melbourne with a public IP that follows your workload wherever it runs.
If uptime matters to your business, we would genuinely like to help you get the design right. Whether you need a single resilient cloud server, a full virtual data centre, or a triple site replicated environment for a workload that simply cannot stop, talk to our team and we will help you build exactly the level of high availability you need on mCloud. After twenty years of engineering everything we do around uptime, keeping Australian businesses online is what we are here for.
Build your own cloud server online and see the price instantly with the mCloud calculator:
micron21.com/calculator
Configure exactly the CPU, RAM, storage tier and redundancy level you need, compare it against AWS, Azure and the other Australian providers, and see for yourself how much you can save by running on a platform we own end to end. There is no sales call required to get a price. Build it, price it, and when you are ready, deploy it.
