Telstra outage traced to NTP server stuck in 2006

Telstra has disclosed that the network failure behind a major Australian mobile outage was triggered by an NTP server that came back online with its clock set to 2006.

The incident caused widespread connectivity problems across Telstra’s mobile network, including outages affecting Australia’s 000 emergency services line, as well as disruptions to electronic payments services and transport networks. Telstra detailed the cause in a submission to a Senate inquiry examining outages that affect emergency services. That inquiry initially focused on rival Optus, whose own outage has been linked to multiple deaths after people were unable to reach emergency services.

According to Telstra, the chain of events began during work to fix a known resilience issue: a faulty backup power feed in the chassis housing one of its NTP servers. A technician began replacing the chassis a few minutes before midnight on July 7, and had completed the work and powered the server back on by 3:38 AM on July 8.

When the server restarted, its GPS card “did not operate as expected,” Telstra said. The company now believes that happened because of an intentional design change made earlier to fix another fault, but that change had not been properly documented.

Recommended reading

Railway parts robot lifts VLMZ welding output 10%

“This meant the maintenance team was not aware of the way the device would behave when restarted.”

Telstra also admitted it had not applied a software update to the device, despite knowing it was available in early 2026.

“Had that software update been completed or had the design change been properly reviewed and documented post the earlier incident, and reflected in the maintenance procedure, the outage may not have occurred.”

Once online, the server distributed the wrong time across the network. Other systems then checked digital certificates, concluded something was wrong, and refused connections. Only one of Telstra’s three NTP servers was affected, but that was enough to create broad disruption.

Telstra isolated the faulty server at 7:11 AM, and by 10:30 AM had identified all network components that had taken the bad time data. Recovery was not immediate: as correct time propagated through the network, some equipment failed to close IP sessions, meaning some customer devices could not reconnect unless they were rebooted.

Telstra called the outage “clearly unacceptable” and said its investigation will examine why the design change was not documented, why the software update was missed, and what controls need to change so known risks are captured and closed before they hit customers.