Active - Storage latency, timeouts, or HTTP 500 errors in South Central US

Starting at 18:44 on 26 December 2024, a power incident in a single availability zone in the South Central US region impacted the availability of multiple Azure services.Impacted ServicesVirtual Machines: RecoveringAzure Cosmos DB: Recovering Current Status: We identified that the incident was caused by a power failure in a single availability zone in the South Central US region, which led to a temporary loss of service availability for several Azure services with dependencies on the impacted infrastructure. The underlying power issue was identified and then isolated to allow the process of power restoration to begin. Power has been restored at 20:43 UTC on 26 December 2024. The structured service recovery process mitigated the issue for the majority of affected services. During this incident, we also identified two related issues which impacted a set of dependent services, this included: impact to a set of network devices in the region; and, impact to a networking service resulted in timeouts to a set of Virtual Machines.Timeline:18:44 UTC on 26 December 2024 - Incident auto-detected with cause identified20:30 UTC on 26 December 2024 - Power issue identified and isolated.20:43 UTC on 26 December 2024 - Power has been restored. Service recovery process underway22:30 UTC on 26 December 2024 - Set of network devices impacted due to incident. Some dependent services recovering00:30 UTC on 27 December 2024 - Network devices continue recovery02:30 UTC on 27 December 2024 - Network device recovered; additional, related, networking issue identified impacting dependent services08:30 UTC on 27 December 2024 - Ongoing mitigation of additionally impacted services13:00 UTC on 27 December 2024 - Mitigation to most affected servicesIf your services are impacted, and you have a pre-configured failover option for a different Availability Zone or Region, we recommend evaluating this option until your services are fully restored.