Emergency maintenance on a compute node: some virtual machines will be interrupted

The maintenance is complete. The node has been rebooted and the affected virtual machines are back online. If you notice any unexpected behavior, please contact our support team.

We apologize for the inconvenience.

Following a storage incident detected today on one of our compute nodes, we need to reboot this node to restore normal operation.

Expected interruption: 20 to 30 minutes

Impact: virtual machines hosted on this node will be stopped during the maintenance and restarted automatically afterwards. Because of the incident, these virtual machines cannot be live-migrated to another node beforehand, so a short downtime is unavoidable. GPU-enabled virtual machines are affected. Stored data is not affected.

We will post an update once the maintenance is complete.