The Optus outage cannot be considered a “rare occasion”. Over the past few years, we have witnessed several major outages across the telco networks, making it imperative for us to prepare ourselves for such events. But we have ignored the warnings.

Today, more than 99 per cent of telecoms traffic comprises data. Virtually every organisation and nearly all Australians rely on data services through their phones and fixed-line connections. As we’ve observed, an outage of this magnitude can cause significant disruptions in the economy and people’s private lives. In this case, even the 000-emergency service on landlines was disconnected.

Ranish Kromodoyo and Anya Absalom struggle to make a call outside an Optus store.Credit: Chris Hopkins

These outages are of national interest, and thus, we require national solutions to mitigate the considerable fallout from such events.

What occurred at Optus was likely a software problem. While such issues occur more frequently, most systems recover in seconds or minutes, resulting in minimal disruption. However, in some cases, as appears to have happened this time, a critical fault during a software update can cascade through the computer systems that underpin the network’s operation.

Unravelling, fixing and bringing all these different systems back online can take hours, and sometimes even days. Moreover, not all systems are likely to come back online simultaneously; they need to be restarted one by one, further extending the recovery time.

In the end, this is an infrastructure problem.

There are essentially two long-term solutions. The first pertains to the individual networks of the operators. It is unacceptable for there to be a single point of failure in a network that can bring down an entire country, or as seen before, the entire east coast. With more than 100 years of telecoms experience and a wealth of engineering knowledge and skills, networks can be designed to eliminate single points of failure.

Customers at the Westfield Penrith Optus store during the outage on Wednesday.Credit: Wolter Peeters

In the event of a disruption, traffic should be rerouted through other network systems. In other words, there should be duplicated, unconnected systems, whereby one can take over from the other in emergencies.