Starting at about 2pm today, QRIScloud has been experiencing problems in the NeCTAR / OpenStack control infrastructure. This may be causing NeCTAR Dashboard to not work properly, and OpenStack API calls may be failing as well.
There are indications that NCI is seeing similar issues.
There are no signs that this is impacting on users' existing VMs. However, if applications in your VM are trying to use OpenStack APIs (to dynamically launch instances, etcetera) then these applications may be seeing the same problems.
We will post an update when more information becomes available.
UPDATE (5pm): The issues have now been resolved, at least in QRIScloud.
It was an "Oslo messaging" problem, and the solution was to roll forward on the Juno upgrade (stage 3) ... which we were planning to do this week anyway.