Last night, one of the main Z9000 network switches in the QRIScloud networking fabric failed due to a suspected hardware problem. The other switch in the redundant pair has taken over and is working normally. However, this means that networking services for most QRIScloud NeCTAR instances are currently running without any redundancy.
A QRIScloud engineer is currently investigating the nature and extent of the hardware issue.
UPDATE 2018-05-17 13:10: The network switch has been restored to service, and the hazard is over.