Hi,
Just to let you know that this morning, some jobs used all the nodes that we still had already in "provisioned" state at the Duffy level, but that after that, no nodes were available to fulfil the other requests.
This is due to a network issue we suffered during the week-end (and already contacted the IT team managing the switches to analyze the root cause), and because of that, the dispatcher process (in charge of deploying in advance the duffy nodes) put all provisioning steps into failed mode.
I updated that back to normal, and dispatcher started to reinstall nodes and put those back into the "Ready" pool.
Sorry for the inconveniences