Datacenter issue impacting network

Hello,

In order to protect the equipment, we decided to stop preventively about 60% of the worker nodes in the batch system.

In the meantime, a technician was dispatched and fixed the cooling, we have confirmed that the overwhole temperature is back within sensible values. The worker nodes will be rebooted tomorrow morning to provide the whole capacity to the batch system

Cheers, Romain

Hello,

There seems to be an issue in the datacenter, as the temperature of all servers are off-the-chart. In particular, the NAT which allows you to access public IPv4 addresses from the batch system keeps stopping to protect itself.

Investigation is under way.

Cheers, The T2B IT Team