Back to the archive

The encyclopedia · Software & IT · Technical decision · 2022

KDDI's 61-hour outage hits millions across Japan, the carrier's biggest failure

A software defect during a routine routing-data update took down Japan's #2 carrier for 61 hours and forced ¥7.3 billion in compensation.

KDDI · 2022-07-02

What happened

At 1:35 AM on July 2 2022, KDDI, Japan's second-largest mobile carrier, suffered a massive service outage that lasted about 61 hours, through to 3 PM on July 4. Voice and data services failed across the country, disrupting millions of customers at the start of a long weekend. The ministry report called it 'a serious accident,' and it was the biggest failure in the carrier's history.

The cause was a software defect that surfaced while KDDI was updating routing data. The update triggered a fault that cascaded through the network, and the failure of signalling functions meant voice calls and data sessions could not be set up or maintained. Because the problem lay in the core network logic rather than in any single piece of hardware, simply restarting equipment did not restore service, and recovery dragged on for days.

The final ministry report counted more than 30.91 million affected customers: around 23.16 million voice users and more than 7.75 million data users, with earlier estimates peaking near 40 million lines. KDDI's president, Makoto Takahashi, apologized publicly, and the company announced compensation of 200 yen (about $1.51) to each of its 35.89 million customers, plus two days of basic fees for voice-only subscribers — a total outlay of roughly 7.3 billion yen.

The outage was a decision failure twice over: the routing-data update was deployed without catching the software defect behind it, and once triggered, the failure cascaded instead of being contained. KDDI's compensation bill, the scale of the disruption and the damage to a national carrier's reliability all flowed from a routine update that went wrong at national scale.

Why it happened

  • A software defect inside a routine routing-data update took down core network functions, and that defect went undetected before deployment
  • The failure cascaded through the network rather than being contained, so voice and data stayed down for 61 hours instead of minutes
  • With no rapid recovery path, KDDI had to pay 200 yen to each of 35.89 million customers, a roughly 7.3 billion yen compensation bill
  • The outage hit at the start of a long weekend, maximizing the disruption to a national carrier's users
What it cost¥7.3B to 35.89M customers; 30.91M+ affected for 61hcostly

The lesson

A national carrier's core network is a single failure domain: a defect in one update can take down voice and data for millions for days, so deeper systems need more testing and faster rollback.

Sources

spotted an error? The club wants to know.

Comments · 0

    Sign in to join the comments.

    More like this

    Somewhere, someone solved the problem this company failed at. 2nd Opinion →