Some Atlas users may be unable to access Charts and the Visualization page due to repeated login prompts or unauthorized errors when attempting to sign in. Our engineers have identified a recent change as the likely cause, have rolled it back, and are monitoring the service to confirm that normal access has been restored.
We are no longer seeing capacity issues in Azure East US 2. This issue is now resolved.
We are no longer seeing intermittent spikes in HTTP 503 errors.
This incident has been resolved.
This incident has been resolved.
This incident has been resolved.
Azure has an active service issue for virtual machines in East US 2. According to their service health page, "customers using Virtual Machines in East US 2 may have encountered errors when performing service management operations - such as create, update, scale, or start - for resources in this region. [...] During peak impact, an unexpected surge in usage caused backend Virtual Machine components to reach operational limits, resulting in deployment failures for some VMs. We are actively mitigating through load balancing and resource reclamation. Capacity constraints for affected SKUs persist, and our teams continue working to restore normal levels."
We are actively engaging with Azure support.
At approximately 10:30 AM CT, we identified an issue affecting stream processing within the AWS US-EAST-1 region. As of now, normal operations have resumed. We have initiated a full investigation to determine the root cause and prevent future recurrence.
Users may have seen tools and integrations that rely on older MongoDB Atlas Admin API versions (for example, 2023-01-01 and 2023-02-01) start failing with HTTP 410 responses indicating that the requested version was no longer available. This affected workflows using the MongoDB Atlas Admin API through the Terraform provider, the Atlas CLI, and custom automation.
This behavior was triggered by our planned retirement process for older API versions, but the resulting user impact was unintended because these versions were still in active use.
We have extended the end-of-life date for the affected API versions to mid-2027 (tentative date 2027-07-01) and deployed a configuration change to restore compatibility for these workflows. All affected tools and integrations should now be operating normally & will be updated prior to the modified end-of-life-date.
From 2026-03-01 00:00 UTC to 2026-03-02 12:01 UTC, users relying on these older API versions may have experienced failed requests and associated automation errors.
No user action is required. Previously failed operations can be retried and should now succeed.
This incident has been resolved.