
Three major service disruptions in five days have raised fresh questions about centralized infrastructure risk. On Thursday, Nov. 12, a Microsoft outage interrupted workplace services worldwide. The following day, a telecommunications failure at the Federal Aviation Administration prompted a nationwide ground stop. Now, on Monday, customers of several major U.S. banks are reporting problems accessing their accounts and completing transactions.
The incidents have not been linked, but each illustrates how dependent businesses and the public have become on digital systems they cannot independently control.
Microsoft’s Outage Spread Across Its Global Network
Microsoft’s outage affected Exchange Online and several Microsoft 365 services, including features of Teams, SharePoint and OneDrive. Email access was particularly affected, leaving users unable to reach their mailboxes and delaying message delivery into the weekend.
In a preliminary report released Nov. 15, Microsoft identified a software defect introduced months earlier that remained undetected until changing network conditions triggered it.
The company said an increase in traffic to a shared networking gateway, combined with temporarily reduced server capacity, caused servers to become overwhelmed. As affected servers stopped responding, Microsoft’s automated systems redirected traffic to other available servers. Those servers encountered the same defect, causing the problem to spread.
At the height of the incident, more than 80% of the affected gateway servers were unresponsive. Microsoft also acknowledged that protections intended to isolate failures between parts of its infrastructure did not prevent the disruption from spreading globally.
Restoration required engineers to correct the defective software and gradually bring servers back online. Microsoft reported that active customer impact ended Friday, Nov. 13, and said additional findings would follow.
The incident illustrates a difficult problem in centralized cloud infrastructure. Redundancy can protect against isolated equipment failures, but systems sharing the same software or network dependencies may remain vulnerable to a common failure. In Microsoft’s case, the mechanisms intended to maintain availability became part of the cascading disruption.
FAA Network Failure Disrupted Air Travel Nationwide
On Friday, Nov. 13, the FAA issued a nationwide ground stop after instability in its telecommunications network affected communications and radar information used by air traffic controllers.
A detailed FAA statement released Sunday identified excessive network traffic as the immediate problem. The agency reported approximately 1,000 connection attempts every four to five seconds, overwhelming parts of its aging telecommunications infrastructure.
Investigators identified two possible sources: a radar facility in Amarillo, Texas, and a radio communications facility in El Dorado, Arkansas. Contractor L3Harris had been performing routine network work at the Arkansas location, including installation of a new router.
The resulting instability affected 17 of the FAA’s 22 regional air traffic control centers and spread to additional facilities. The agency said the network stabilized after the suspected locations were isolated. As a precaution, similar functionality was disabled on approximately 60 other routers.
Air traffic resumed Friday evening after a ground stop lasting nearly five hours. Approximately 3,500 aircraft already in flight landed safely. The FAA reported about 12,462 flight cancellations that day, although it cautioned that not all were directly attributable to the ground stop.
Officials said there was no indication of a cyberattack. The precise root cause remains under investigation.
The FAA’s response was a necessary safety measure, and its procedures allowed aircraft already in flight to land safely. But the disruption demonstrated how instability originating at a small number of network locations could affect an entire national transportation system.
Banking Disruptions Add to the Week’s Outages
On Monday, Nov. 16, customers of several major U.S. banks reported problems accessing financial services. JPMorgan Chase experienced widespread online banking disruptions, with thousands of customers reporting difficulty logging in, viewing account balances and completing transactions. Chase acknowledged the problems and said it was working to restore service.
Customers of Bank of America and Wells Fargo also reported disruptions, although the extent of those problems remained unclear as of Monday afternoon. The cause of the banking issues had not been established, and there was no confirmed connection to the Microsoft or FAA incidents.
The banking reports add another example of how dependent everyday activities have become on systems outside users’ control. When digital banking services become unavailable, customers may lose access to basic account functions even though the underlying financial institutions remain operational.
Why Redundancy Does Not Always Prevent Cascading Failures
The Microsoft and FAA incidents involved different technologies and operational requirements. Both, however, exposed the challenge of containing failures within highly interconnected networks.
Microsoft had redundant systems and automated recovery mechanisms. The FAA maintained backup communications capabilities and established safety procedures. Those protections mattered, but they did not prevent widespread disruption.
The distinction is important. Centralized infrastructure risk does not necessarily mean that a system lacks backups or alternative connections. It can also arise when otherwise separate components depend on common software, routing mechanisms or network management systems.
We examined a related concern in our Oct. 30 article, Centralized Exchange Risk After the Binance Hack, which looked at how compromised internal systems can put customer assets at risk even when private keys remain secure. The circumstances differ, but the underlying concern is similar: safeguards can fail when too much depends on a shared system or operator.
Microsoft’s preliminary findings are especially revealing. When traffic was redirected away from failed servers, the additional load helped overwhelm the remaining servers. The infrastructure was distributed across multiple regions, but the shared defect allowed the problem to spread.
For organizations that depend on large technology providers, this raises a practical question: How independent are the systems intended to provide redundancy?
Having multiple servers or facilities offers limited protection when the same underlying problem can affect all of them.
Where Decentralized Infrastructure Fits
Decentralized infrastructure offers another approach to managing some of these dependencies. Rather than concentrating computing resources or network services under one operator, distributed networks can draw on independently operated infrastructure.
Akash Network, for example, operates a decentralized marketplace where computing capacity is supplied by independent providers. Manifest Network is developing decentralized networking infrastructure intended to connect resources across distributed environments.
These networks are not direct substitutes for Microsoft’s enterprise applications or the FAA’s air traffic control systems. Their relevance is in how infrastructure can be organized to reduce dependence on a single operator or operating environment.
A distributed computing network may offer alternative capacity when one provider becomes unavailable. Independent network paths can also help maintain connectivity when individual components fail.
Neither benefit is automatic. Distributed systems can still share vulnerable software, routing dependencies or authentication services. Their resilience depends on whether those dependencies are genuinely independent and whether workloads can move between providers without creating new bottlenecks.
The objective is not simply to increase the number of participating computers. It is to prevent failures in one part of a network from unnecessarily disabling the rest.
The Cost of Shared Dependencies
The Microsoft and FAA failures, followed by Monday’s banking disruptions, have affected very different parts of everyday life. The technical causes may differ, but the consequences share a familiar characteristic: when essential systems become unavailable, the people and organizations that depend on them have limited alternatives.
Decentralized infrastructure is not immune to failure, but reducing dependence on common systems and individual operators can limit how far those failures spread. That principle matters as more business operations and public services rely on increasingly interconnected networks.
Disclosures: This article is for informational purposes only and should not be considered financial, legal, tax, or investment advice. It provides general information on cryptocurrency without accounting for individual circumstances. Sarson Funds, Inc. does not offer legal, tax, or accounting advice. Readers should consult qualified professionals before making any financial decisions. Cryptocurrency investments are volatile and carry significant risk, including potential loss of principal. Past performance is not indicative of future results. The views expressed are those of the author and do not necessarily reflect those of Sarson Funds, Inc. By using this information, you agree that Sarson Funds, Inc. is not liable for any losses or damages resulting from its use.







