On 1 October 2026, between approximately 08:00 and 13:47 UTC, the Mosaic platform was intermittently unavailable due to a network failure at one of our infrastructure providers. During this period, there were several outages in which the platform was fully unreachable.

We moved the platform to our standby environment and Mosaic has been running stably since.

This is a preliminary report. We will share a detailed report, including a full timeline and confirmed preventive actions, once our internal review is complete.

During the outages, management portals, service APIs, and authentication endpoints did not respond. Requests timed out or returned connection errors.

Managed services were restored first, followed by customizable services. A small number of customer-specific setups needed additional steps afterward, such as DNS updates, which we handled individually with the affected customers.

The impact varied depending on the services used and any customer-specific setups. The Mosaic database was not affected and no data was lost.

Axinom DRM was affected by the same provider issue to a much smaller extent: less than 1% of license requests in the Europe region failed intermittently until we removed the affected provider from the routing. We are sharing a separate report for this.

Our infrastructure provider had a failure in its data center network (a virtual switch failure affecting private networking and load balancers). The provider has since confirmed this cause. It broke the load balancers in front of Mosaic and the communication between the servers inside our cluster, so the platform could not serve requests. The issue was not caused by a defect in Mosaic, a deployment, or any change on our side.

The recovery took longer than expected because the provider’s network failure was intermittent, with service recovering temporarily before failing again. This made it difficult to determine whether a switchover was necessary. When the issue continued to recur, we proceeded with the switchover.

We eventually restored the services by completing the move to our standby environment, which is operated by a different cloud provider. The platform has been running stable there since.

We are currently completing our analysis of the incident and deciding on the measures we will take to prevent a comparable outage in the future. We will provide a full incident report later this week. It will include a detailed timeline, the root cause analysis and these measures.

Update (9 October 2026): The full incident report will now be shared next week, as our analysis is taking longer than expected.

Update (15 October 2026): The full incident report includes details about our internal infrastructure that we don’t publish openly, so we have shared it by email with the affected customers instead.

We understand that an outage of this length is not acceptable, and we sincerely apologize for the impact on your work. If you have questions about this report, please contact us at support.axinom.com.