Chapter 1: From the Telstra Outage to the Optus Outage

In our previous two-part series, we examined a nationwide outage at Telstra, one of Australia’s largest telecommunications providers. On July 8, 2026, maintenance work on servers responsible for network time synchronization triggered a nationwide outage affecting approximately 8.8 million customers, which also disrupted connections to the emergency number Triple Zero (000). We concluded that the root cause was not the time drift itself, but gaps in operational management: an undocumented design change and an update that had never been applied.

At the end of Part 2, we noted that the incident followed the Optus emergency call outage of September 2025, and that not only individual companies but the telecommunications industry as a whole was still in the middle of continual improvement.

About two months later, on September 11, 2026, Optus experienced another outage, and calls to 000 were affected. 000 is a single number that combines what in Japan are two separate numbers: 110 (police) and 119 (fire and ambulance). Victorian Health Minister Ingrid Stitt stated that this was the third outage at Optus or Telstra in the past 12 months.

Last time, we focused on change management within a single company. This time, we broaden the perspective. Within 000, a system supported by multiple carriers, who protects availability, and how? We explore this question through the lens of ITIL availability management.

Chapter 2: How a 000 Call Flows: A Single Value Stream and “Mutual Backup”

To understand this outage, we first need to understand how 000 is structured.

In Japan, the number you call depends on your situation: 110 for police, and 119 for fire and ambulance. In Australia, by contrast, people call a single number, 000, whether they need police, fire, or ambulance. From a mobile phone, calling 112 also connects to the same emergency call service as 000.

The mechanism also differs significantly. When someone calls 000, they first speak with a Telstra emergency call operator, who then transfers the call to the appropriate agency: police, fire, or ambulance. Unlike 110 or 119 in Japan, which connect directly to police or fire services, the first point of contact is a reception desk staffed by a telecom operator. Telstra has played this role since 1961 and answers more than 32,000 calls to 000 every day. In other words, regardless of which carrier a caller uses, every 000 call ultimately reaches Telstra’s reception.

Another important element is how mobile connections work. When a mobile user calls 000, the phone first tries to connect through their own carrier’s network. If that network is unavailable, it may automatically try to connect through another carrier’s network.

With this structure in mind, the relationship between the previous outage and the current one becomes clear.

In the earlier Telstra outage, the network of the very company responsible for 000 reception went down. In that incident, 604 calls to 000 failed to connect. In the latest Optus outage, calls were stopped at the Optus network stage, before they could reach Telstra’s reception.

The points of failure were different. But from the user’s perspective, both were the same service failure: 000 could not be reached. 000 is not a service that any one company delivers on its own. It can be seen as a single value stream linking multiple carriers’ networks, Telstra’s reception, and the emergency services of each state. And each carrier’s network also serves as a backup route for the others.

Chapter 3: Through the Lens of Availability Management: What Your Own Availability Cannot Protect

In ITIL, the purpose of availability management is “to ensure that services deliver agreed levels of availability to meet the needs of customers and users.” Availability here refers to “the ability of an IT service or other configuration item to perform its agreed function when required.”

Key metrics for availability management include the number of service interruptions, total downtime over a given period, and mean time to restore service. It is of course important for each carrier to measure and improve these metrics for its own network.

However, what 000 users need is not “my carrier’s network is up.” It is “I can reliably reach 000 in an emergency.” This is the availability of an entire value stream spanning multiple organizations, and it arguably cannot be fully captured by any single company’s metrics.

The “backing each other up” structure also rests on certain preconditions. Rerouting to another carrier’s network only works if that network is healthy and can absorb a sudden surge in traffic. In fact, Telstra has previously told the ABC that roaming onto other networks during a major outage could place an extraordinary load on the remaining networks and could itself cause problems. Having a backup route and having one that actually works are two different things, and this also raises questions of capacity and performance management.

In the basic terms of ITIL 4, 000 is not a service provided by one company. It is a service made possible through value co-creation among multiple service providers and government agencies. One of the recommendations for availability management is to integrate the practice into the organization’s value streams. For a system like 000, that integration has to extend beyond a single organization and across organizational boundaries.

Chapter 4: The Senate Report: A Push for Protection by Design, and the Open Questions

This issue has also been debated in Australian politics. On September 21, 2026, a Senate committee inquiry into the Triple Zero crisis released its final report. The inquiry was launched in response to the September 2025 Optus outage and was later expanded to cover other call failures that came to light.

The report made 11 recommendations. Key ones include mandating domestic roaming, introducing minimum performance and reliability standards with automatic compensation during outages, and enabling people to send text messages to 000 in emergencies. The most significant recommendation is to transfer the work of answering 000 calls and forwarding them to local emergency services, currently handled by Telstra, to a newly established national body.

Positions differ, however. Telstra has signaled willingness to work with the government and others on performance standards, but maintains its opposition to mandatory roaming, out of concern that a mandate would undermine infrastructure diversity and resilience over the long term. Labor members of the committee supported most of the recommendations but expressed reservations about mandatory roaming and some others, arguing that calling 000 through another network is already possible and that roaming is not necessarily the best way to improve connectivity. The ACCC (Australian Competition and Consumer Commission) rejected mandatory roaming in 2017 but is now reconsidering it as part of a new inquiry.

The government said it would consider the recommendations, pointing to steps it has already taken: raising the maximum penalty for breaches to A$30 million, introducing an outage reporting mechanism and a public register, and establishing the Triple Zero Custodian to oversee the system.

How should a structure that concentrates reception in a single company be handled? Should mutual backup between carriers be codified as explicit rules rather than left to custom? From an ITIL perspective, these debates are fundamentally about how to design availability management and IT service continuity management across organizations.

Chapter 5: Bringing It Home: Checkpoints for Cross-Organizational Dependencies

Most corporate systems do not operate at the scale of a national emergency call system. Still, the underlying structure is the same: your services rely on other companies’ services, such as cloud platforms, network lines, and external APIs. Consider checking the following points.

First, what scope do you measure availability over? Beyond your own system’s uptime, are you looking at whether the service is actually usable across the entire value stream, as users experience it?

Second, the assumptions behind your backups. Is your “switch to another route during an outage” design backed by an agreement with the other party? Can the failover destination actually handle the load?

Third, communication during an outage. Have you decided who informs related companies, regulators, and users, what they say, and in what order? The outage communication design we discussed in Part 2 is truly put to the test in cross-organizational situations.

Conclusion: Availability Cannot Be Achieved Within a Single Company

Last time, we looked at a case where poorly managed change within one company led to a nationwide outage. The latest Optus outage and the Senate report show that even when each company improves internally, the services users rely on are built across organizational boundaries.

The cause of this outage has not yet been disclosed in detail. Key things to watch are the findings of the Triple Zero Custodian and ACMA investigations, and the government’s formal response to the Senate recommendations. We plan to revisit this topic as further developments emerge.

Background Information

Telstra: The telecommunications company with the largest share of Australia’s retail mobile market. Since 1961, dating back to its predecessor Telecom Australia, it has handled first-response reception for 000 emergency calls. The July 8 outage covered in our previous article affected approximately 8.8 million customers.

Optus: Australia’s second-largest telecommunications company, owned since 2001 by Singapore’s Singtel. It has approximately 11 million mobile subscriptions and about 1.3 million broadband services, and its national mobile network is second in scale only to Telstra’s. It experienced an emergency call outage in September 2025, over which ACMA has taken Optus to the Federal Court.

References

ABC News (2026). “Victoria Police conduct dozens of welfare checks following Optus outage.” https://www.abc.net.au/news/2026-09-12/optus-outage-emergency-services/107146350
ABC News (2026). “Triple Zero should be taken out of hands of major telcos, Senate inquiry recommends.” https://www.abc.net.au/news/2026-09-21/triple-zero-senate-inquiry-recommendations/107162636
Telstra. “Triple Zero (000) and 112 Emergency Call Service.” https://www.telstra.com.au/consumer-advice/emergency-call-service
Information Age (ACS) (2026). “Another Optus mobile outage blocks 46 Triple Zero calls.” https://ia.acs.org.au/article/2026/another-optus-mobile-outage-blocks-46-triple-zero-calls.html
Information Age (ACS) (2026). “Singtel prepares Optus sell-off.” https://ia.acs.org.au/article/2026/singtel-prepares-optus-sell-off.html
Information Age (ACS) (2026). “Triple Zero should be taken out of Telstra’s hands.” https://ia.acs.org.au/article/2026/triple-zero-should-be-taken-out-of-telstra-s-hands.html
IncidentTech Inc. (2026). “Why Did Maintenance Work Cause a Nationwide Outage? [Part 1]” and “Outages Happen Even When Procedures Are Followed [Part 2]”