CyrusOne Blames CME Data Centre Outage on Human Error

On November 28, 2025, a CME Group data centre outage occurred due to a human error at a CyrusOne facility. Specifically, on-site staff failed to follow customary procedures for draining cooling towers prior to freezing temperatures. The servers that hosted the CME’s Globex trading platform overheated and shut down automatically to minimise damage as a result of the cooling system failure, which resulted in a trading standstill lasting more than ten hours at the end of last week.

A representative for CyrusOne, a business owned by KKR & Co. and Global Infrastructure Partners, told Bloomberg News on Saturday that onsite employees and contractors at the Aurora, Illinois facility disregarded standard protocols for emptying cooling towers prior to freezing temperatures. As a result, the cooling system became overworked and the temperature increased.

The CME, one of the biggest derivatives exchanges in the world, uses the data centre as its central location to handle trillions of dollars every day in the global equity, currency, bond, and commodity markets.

On November 28, a technical issue at a facility about 50 miles (80 kilometres) from Chicago affected markets from Tokyo to London, stopping trading in everything from gold to oil and interest rates.

The data centre’s “initial remediation attempts further exacerbated the problem, which eventually led to the failures of a number of chillers,” according to a separate statement from CME, while CyrusOne claimed that it took “extensive and decisive steps to restore the cooling systems.”

The CME’s dependence on the data centre, which it sold to CyrusOne in 2016 and leased back for 15 years, was highlighted by the outage.

Despite CME’s disaster recovery plan’s recommendation to relocate to a data centre in the New York region, the exchange decided against doing so because the data it had indicated a temporary outage.

CME declared on Saturday, “We acknowledge the impact experienced by our clients worldwide.”

In order to manage the outage and its consequences, CyrusOne claimed to have sent more staff to the site. Executives like Chief Executive Officer Eric Schwartz and elite engineering teams were among them. In order to restore cooling capacity, it additionally employed extra equipment. The facility never lost electricity, according to the data centre operator.

The company added that in order to avoid a recurrence of the outage, it modified its cold weather protocols, increasing the number of engineers on site, hardening its infrastructure gradually, and improving employee training.

For investors who have invested hundreds of billions of dollars in the artificial intelligence (AI) data centre boom, the equipment breakdown has become a wake-up call, reminding them that the facilities’ leasing agreements are, in fact, breakable. Many of them include built-in termination clauses in case of recurrent disruptions.

Goldman Sachs Group Inc. had to postpone a US$1.3 billion (RM5.34 billion) mortgage-bond offering for CyrusOne following the disruption. As previously reported by Bloomberg, the bank was overseeing what would have been the data centre operator’s biggest-ever offering of commercial mortgage bonds, with early comments due on November 25.

Citigroup Inc. emphasised in a study released on Friday that data centre leases may contain provisions that let tenants to end their agreements in the event of specific shortcomings, such as cooling problems.

Citi stated in the report, “We anticipate that this is an isolated incident, and the owner-operator and tenant will fully resolve any redundancy deficiencies that caused the problem.”

The cause of the outage as said was that the contractors and on-site employees at the CyrusOne data centre in Aurora, Illinois, failed to prepare the cooling towers for cold weather according to established procedures. Consequently, the interior cooling system malfunctioned as the outside temperature dropped, causing the server rooms’ temperatures to rise and ultimately triggering the automated shutdown of vital servers.

The market disruption as a result of the outage on November 28, 2025, stopped trading and froze price discovery for a variety of international financial instruments, including futures and options for stocks, commodities (such as gold and oil), currencies, and U.S. Treasuries, which are vital to traders all over the world.

And the contingency plans set in place, due to early indications that the disruption would be brief, CME did not activate its backup facility.

While the Stock Market reaction for the CME Group stock (CME) fell after the downtime, from roughly $281.46 on November 28 to $270.42 by the close on December 8.

However CyrusOne actions in reaction to the incident, CyrusOne said it has revised its cold weather protocols, which include boosting the number of workers on duty during severe weather and improving cooling infrastructure to avoid repeat occurrences of this kind of failure.

And the business impact on the incident caused Goldman Sachs to postpone CyrusOne’s planned $1.3 billion mortgage-bond offering and brought attention to the fact that data centre leases may include provisions that permit tenants to end agreements in the case of serious service problems.