Key points
- Microsoft is in the midst of a major 12 hour-plus outage affecting a number of M365 services, notably Exchange Online access, worldwide.
- The vendor traced the problem to a specific component on affected servers that processes health checks, becoming overwhelmed and causing capacity loss.
- Microsoft is testing a second potential recovery method and shifting traffic to alternate routes, with users possibly experiencing a gradual recovery as capacity is restored.
Microsoft is in the midst of a major 12 hour-plus outage impacting a number of M365 services, notably Exchange Online access, worldwide.
Australian enterprise users awoke to email not loading and an error message that “something went wrong” with a “recommended action” to refresh.
The vendor first acknowledged the issues at 4.54am AEDT.
The company’s status page ndicated that Teams, SharePoint Online, Microsoft Copilot 365 and Microsoft Defender XDR could be impacted.
A broader selection of services including Intune, Purview and Power BI have since been listed as being affected.
As to the cause, Microsoft said a “portion” of its server infrastructure is “unexpectedly unavailable; straining the remaining infrastructure.”
It identified a “specific component” on the servers that processes health checks as the problem.
“We found a core process on affected servers is becoming overwhelmed, causing capacity loss,” it said.
Initial attempts to bypass that component with a “configuration change” yielded results – allowing “a subset of affected servers to successfully return to service,” Microsoft said just after 8am AEDT.
However, the bypass did not ultimately lead to the recovery of services.
Microsoft then isolated a portion of its infrastructure to validate a second “potential recovery” method, but this has also so far proven unreliable.
“The servers we returned to service can become unresponsive again once they receive traffic; we’re analysing diagnostic data to determine what’s triggering the behaviour,” Microsoft said.
“In parallel, we’re temporarily limiting which types of traffic can reach the affected infrastructure in some locations.
“Additionally, we’re rebuilding a subset of affected servers using a previous configuration to validate if this prevents them from becoming unresponsive.”
Microsoft has so far not indicated in which cloud region the impacted servers are located.
It added on X that it is “aggressively pursuing all pathways to reduce impact”, including an internal pause on “all deployments”.
