Microsoft's Outlook outage post-mortem reveals testing gaps and failover backfires
Microsoft has published its final report on the Outlook outage, which stemmed from a July update to gateway servers handling Microsoft 365 connections. Load tests ran on older, lower-core machines, so the defect went undetected on the newer high-core servers where it actually appeared. Safeguards such as automatic failover and traffic management then spread the load across healthy servers, which were overwhelmed as recovering servers came back online. Within roughly two hours, every region was affected. Of 20 planned repair items, seven are complete, with 13 still pending, including testing on the newest hardware.
This is an AI-generated summary. ShortSingh links to the original source for the complete article.

Discussion (0)
Log in to join the discussion and vote.
Log in