[osc-secnum-fr1] Network disruption to access region applications

Update timeline

1. postmortem Aug 25, 2026, 12:44 PM UTC

   During the evening of June 24, starting at approximately 21:30, Scalingo experienced significant disruptions affecting both the osc-secnum-fr1 and osc-fr1 regions. The incident occurred during an extreme heatwave affecting the area where one of our infrastructure provider's datacenters is located. Under these conditions, cooling systems at the datacenter began to fail, resulting in severe overheating and impacting some of the provider's infrastructure. Although Scalingo's infrastructure is not directly hosted in the affected datacenter, the incident had cascading effects on some services we rely on: * **osc-secnum-fr1:** A critical network component operated by our infrastructure provider was hosted in the affected datacenter, resulting in network unavailability towards the region. Connectivity was restored at approximately 23:10, after which applications and databases in the region became reachable again. We observed no additional impact resulting from the incident after connectivity was restored. * **osc-fr1:** Although object storage data is replicated across multiple sites, we identified that our provider's API used to access the service was hosted exclusively in the affected zone. This resulted in minor disruptions to operations relying on object storage, including some deployment operations.