[Critical] Issues with All Files Page, Box Notes, API calls, logins and downloads
Final update
We recently addressed an issue that briefly affected the All Files page, Box Notes, API traffic, logins, and downloads in Box. We would like to take the opportunity to further explain this issue and the steps we have taken to keep it from happening in the future. On Tuesday, September 29, 2026 between approximately 6:39 AM PDT and 6:44 AM PDT, some customers may have experienced errors when using Box. The issue occurred when several database instances simultaneously became unavailable due to a disruption in our cloud-managed infrastructure. Our internal systems detected the unhealthy instances and automatically rerouted traffic to a healthy copy, which fully restored service availability. # Analysis Our database service is generally deployed evenly across multiple availability zones in order to reduce coupling to a single failure domain within our cloud infrastructure. However, VM scheduling within a given availability zone is managed by the compute service itself. In this instance, several VMs within a single availability zone were attached to a shared network device prior to the incident. When that device experienced a brief interruption, multiple database instances simultaneousl
Timeline
- Postmortem · Oct 27, 21:25 UTC
We recently addressed an issue that briefly affected the All Files page, Box Notes, API traffic, logins, and downloads in Box. We would like to take the opportunity to further explain this issue and the steps we have taken to keep it from happening in the future. On Tuesday, September 29, 2026 between approximately 6:39 AM PDT and 6:44 AM PDT, some customers may have experienced errors when using Box. The issue occurred when several database instances simultaneously became unavailable due to a disruption in our cloud-managed infrastructure. Our internal systems detected the unhealthy instances and automatically rerouted traffic to a healthy copy, which fully restored service availability. # Analysis Our database service is generally deployed evenly across multiple availability zones in order to reduce coupling to a single failure domain within our cloud infrastructure. However, VM scheduling within a given availability zone is managed by the compute service itself. In this instance, several VMs within a single availability zone were attached to a shared network device prior to the incident. When that device experienced a brief interruption, multiple database instances simultaneously became unavailable. Our systems retain warmed copies of the database instances to take over in the event of a hardware failure, and continuously monitor the health of the service. Within minutes of the impact start, this system correctly detected the issue and successfully remediated the problem b
- Resolved · Sep 29, 15:30 UTC
From approximately 06.39 AM to 06:44 AM US Pacific time, we observed an issue impacting All Files Page, Box Notes, API calls, logins and downloads. Our systems automatically detected and corrected the underlying issue. There is no current impact and no further updates will be provided here. If you continue to experience any issues, please contact Box Support at https://support.box.com.