An issue affecting the internal message queue service that the Nectar Research Cloud relies on has caused intermittent errors during several outage windows since Monday 12 October. During these periods, a range of Dashboard actions and CLI commands may have failed or responded slowly. Actions involving the Volume (Cinder) service were the most affected, with some impact on compute, network and DNS operations.
We have put mitigations in place and service is largely restored. However, a lingering issue with the message queue system is still causing failures in some services. As a last resort, we are rebuilding the message queue clusters. Each rebuild will cause an outage of up to 15 minutes.
Completed
Significant
Last modification at:
Wed, 14 Oct 2026 11:10p.m. (UTC)
We have completed the rebuild of the messaging system. We are resuming the services progressively.
Update posted
3 weeks ago.
Rebuilding in progress
Update posted
3 weeks ago.