Is signageOS down?

We are currently experiencing an issue with Bulk Provisioning. Devices are currently taking longer than usual to provision in bulk. Our team is investigating, and we will post an update as soon as we have more information. Other platform features are not affected.

We have identified the cause of the Bulk Provisioning issue and are working on a fix. We will post another update when the fix is deployed.

We have implemented the fix. The rollout is being prepared. We will post another update when the fix is deployed.

A fix has been released and the queues have been processed. We are monitoring the situation.

We are detecting an issue in device online status detection system. Device's may show incorrect online status. Devices themselves are not affected.

We found and fixed the issue. Everything is back to normal.

We are currently experiencing multiple reconnects on the legacy device gateway. No devices or their playback are affected.

Team found the issue rooted in extensive memory usage. Deploying a production patch.

Reconnection resolved. Team confirms no impact on devices. Majority of devices already running on new gateway and were not impacted in any way.

We detected degradation of the service that propagates certain policy changes to devices, namely changes in applets, runners and plugins. Expect these to be delivered with a delay to devices. No data is lost.

We are detecting degrated performance for certain actions in the system that may lead to API returning 504. The team is investigating the cause.

When: 27 August 2026, 22:45 UTC to 28 August 2026, 12:30 UTC Affected: signageOS production, EU region. The US region was not affected. Summary During planned maintenance to refresh telemetry configuration across managed devices, the volume of work generated by the refresh saturated an internal processing queue. Unrelated platform operations were delayed for about 33 minutes. The same maintenance also cleared each device's most recent telemetry readings instead of marking them as stale. Online devices re-reported within seconds. Offline and infrequently reporting devices showed no latest-telemetry values until we restored them from durable storage the following midday. No data was permanently lost, no device went offline, and content playback was not interrupted. Customer impact Delayed platform operations, 27 Aug 22:45 to 23:18 UTC. Device restarts, applet and content updates, and configuration changes submitted in this window were accepted normally but took longer than usual to reach devices. The backlog cleared in full by 23:18 UTC. Missing latest-telemetry values, 27 Aug 22:45 UTC to 28 Aug 12:30 UTC. The latest-telemetry endpoints and the matching web console values (temperature, storage, brightness, volume, uptime and similar) returned empty for affected devices. Each device repopulated its own values on its next report; the remainder were restored from durable storage. Not affected: device connectivity and content playback, historical and per-type telemetry, all other APIs and console functionality, and the US region. Detection Both issues were identified by signageOS, not reported by customers. Our alerting picked up the queue saturation within roughly 11 minutes, with the engineer running the maintenance already responding. The missing telemetry values were identified by our team the following morning. What went wrong The maintenance used a per-device operation for a fleet-wide job. Each instruction is cheap on its own, but issued to the whole fleet at once they exhausted the capacity of an internal system that all platform operations share, so unrelated customer actions waited behind them. Separately, the store that holds each device's latest telemetry readings was handled in our code as a temporary cache, when it is in fact the primary source for those values. The refresh cleared it rather than marking the values stale. Because online devices repopulate the store within seconds, the behaviour passed its original testing and only became visible when applied fleet-wide to devices that were offline. What we have done The refresh no longer deletes stored telemetry values; it marks them stale as intended. This fix was deployed on 1 September 2026. All affected data was restored from durable records on 28 August. We have also corrected how this store is named and documented internally and removed the destructive operation from its interface so it cannot be reached by mistake. What we are changing We are replacing the per-device fan-out with a single fleet-wide operation, restricted to signageOS administrators and requiring explicit confirmation, a documented reason, an audit record and a cooldown period. Telemetry test plans now include the offline-device case that earlier testing did not cover. We are also reviewing every remaining code path that touches the same store to confirm none of them can delete stored readings.