In a recent incident, Microsoft faced a data center outage in Australia and has attributed the disruption to a combination of insufficient staff capacity and failed automation. The outage occurred on August 30 and was caused by a utility power sag in Australia’s East region, leading to the shutdown of a subset of cooling units in one of Microsoft’s data centers.
Details of the Outage
As a result of the power sag, the cooling units in the affected data center went offline, causing a significant rise in temperature. This temperature surge triggered an automated shutdown of the data center, impacting crucial services such as computing, networking, and storage.
Staffing Issue
While the cooling units could have been manually restarted, the data center faced a shortage of personnel. Insufficient staff members were available at the time to address the issue promptly. Acknowledging this staffing limitation, Microsoft swiftly took action by temporarily increasing the team size, ensuring an appropriate level of personnel for future incidents.
Improving Automation
Following the outage, Microsoft has recognized the need to enhance its current automation systems for better service restoration during similar incidents. The company is committed to strengthening its automation capabilities to ensure uninterrupted services. Efforts are underway to make the automation systems more resilient to different types of voltage sag events, mitigating the risk of potential shutdowns.
Evaluation Process
In light of the outage, Microsoft is conducting a comprehensive evaluation of its data center infrastructure. The aim is to restructure their systems to prioritize the restart of the highest-load servers and corresponding chillers during outages. This evaluation will facilitate a more efficient recovery process, minimizing disruption and downtime for clients and users.
Previous Outages Faced by Microsoft
This recent outage is not an isolated incident for Microsoft, as the company has experienced multiple service disruptions in the past. In both February and January, Microsoft encountered global outages that led to restricted access to email and Teams, impacting businesses and individuals reliant on these services.
Recognizing the significance of uninterrupted service provision, Microsoft has taken decisive steps to address the staffing issue and improve automation within its data centers. The implementation of a larger team size ensures that sufficient personnel are available to swiftly respond to and resolve incidents. Additionally, the focus on enhancing automation systems will bolster service restoration during unexpected events. By evaluating and restructuring the infrastructure, Microsoft is taking proactive measures to prevent future outages, ensuring seamless access to their services for customers worldwide.