From 18/03 to 28/03 we carried out electrical maintenance at the Gravelines site, originally scheduled for 18/03 to 21/03.
These were published on the status-ovhcloud.com website (https://network.status-ovhcloud.com/incidents/st94xrmdkkxg).

During the initial phase, on Monday 18/03, we switched to the backup power line and despite our usual precautions, this change resulted in the loss of several racks, particularly on GRA1/GRA2/GRA3 datacenters.
On Friday 22/03, we restored the GRA4/GRA2/GRA1 datacenters to their nominal status.
However, operations on the GRA3 data centre were different due to an analysis of the incident on 18/03.
As a result of an electrical equipment failure, some racks were no longer electrically redundant.
To restore service to its nominal mode, the decision was made to switch the racks one after the other to recover the electrical redundancy.
This operation was carried out between Monday 25/03 and 28/03 01:40 UTC.
On 27/01, a handling error occurred which resulted in a rack being shut down, but our teams quickly restarted it.
Finally, on 28/03 at 08:15 UTC, the system returned to its nominal state, with no loss of service.

We acknowledge the impact that this situation has had on your business and offer our sincerest apologies for any inconvenience caused.