I've looked at the failures in the past week and they're all "Lock wait timeout" in Neutron.
However, there are several difference occurrences of the timeout error, and I will therefore file different bugs.
I am a bit gutted that the one we fixed - on updating port status - appears to be back.
I think from a gate stability perspective there is no risk at the moment as 99% of the failures is in the full jobs (which we'll enable soon however for neutron only).
I've looked at the failures in the past week and they're all "Lock wait timeout" in Neutron.
However, there are several difference occurrences of the timeout error, and I will therefore file different bugs.
I am a bit gutted that the one we fixed - on updating port status - appears to be back.
I think from a gate stability perspective there is no risk at the moment as 99% of the failures is in the full jobs (which we'll enable soon however for neutron only).