Microsoft reported a global surge after isolation barriers between network partitions failed to contain the disruptions, which are compounded by a separate issue that prevents regional safeguards from stopping the egress. Teams, OneDrive, SharePoint and Office applications were affected; Some staff services remained operational via a fallback, but Exchange Online could not use the same route. Tests missed the flaw due to different traffic patterns, configuration and older hardware, while the issue was more pronounced on newer, multi-core machines. Recovery was lengthy as restored servers were overwhelmed by backlog requests. Standard flags and features failed to reverse the months-old flaw. Engineers restored the code path by taking groups of servers offline, applying updates, and gradually bringing them back into service. Microsoft apologized and announced that a full review of email would be released later this week. Tag: Breaking GlobalHQ.