Microsoft services disrupted in US, nearly 1 lakh Outlook complaints logged

Companies and government offices around the world depend heavily on services such as Microsoft 365, Outlook and Teams for email, online meetings and everyday digital work. A sudden technical fault in them can halt the work of thousands of users.

Reports emerged on Thursday of a disruption in some Microsoft services in the US. The disruption is said to have also affected some connectors used in the AI services of OpenAI and Anthropic. The episode has revived debate over how far cloud servers can be trusted.

According to Downdetector, a platform that tracks outages, a large number of complaints were logged in the US about Outlook. By about 2 pm, the number of complaints had reached roughly 1 lakh. However, these figures do not mean that every user's service was completely down.

Many organisations rely on Microsoft's services to send email, share important files, hold online meetings and carry out routine office work. A disruption therefore does not stop at individual users; it can also affect the work of companies and government offices.

The question is why such servers go down repeatedly. There could be five major technical reasons.

The first is a faulty configuration change. Running cloud servers involves a wide range of settings and network rules. Engineers use automated tools to apply changes to many servers at once. If a wrong IP address, a mismatch in rules or a faulty setting is applied, services can be affected on a large scale. At times the same error also stops the management tools that fix such faults from working.

The second is a flaw in a software update. Cloud platforms are updated continuously for security fixes, bug fixes and new features. But testing in a controlled environment can differ from real-world conditions. Flaws that testing missed can therefore surface once an update is rolled out widely. Such flaws can slow a service down or shut it down altogether.

The third is a sign-in problem in Microsoft Entra ID. This service verifies a user's identity and handles logins. With it, a user can access Outlook, Teams, SharePoint and other apps through a single login. If this identity-checking system is disrupted, a user cannot log in even with the correct password. As a result, many interconnected services can be affected at once.

The fourth is a fault in a data centre's power supply. Thousands of servers run continuously in a data centre. For this there are arrangements such as the main power supply, battery backup and generators. If the systems that transfer power, the backup systems or the power distribution equipment fail, servers can lose power. The services of that data centre are then likely to be affected.

The fifth is a network routing error. Network routing plays a key role in carrying data from one place to another on the internet. It uses the Border Gateway Protocol (BGP). If wrong routing information spreads, data can go along the wrong path or fail to reach the server it is meant for. This can break connections to websites, apps and cloud services.

If a cloud server goes down, the work of companies using cloud services can be delayed. Online meetings can be cut off midway, sending email can become difficult and reaching important files can be hard. This is why companies using cloud services keep backup arrangements, alternative means of communication and emergency working plans ready.

On the whole, there is no single reason for cloud servers going down. Wrong configuration, a flawed software update, a failure of the login system, power supply problems and routing errors can be the main causes behind it.