The data centre has identified the issue with the ceph disk cluster. Ironically, the ceph system is designed to provide multiple layers of disk redundancy. They are working on resolving the issue. The disk I/O wait time is going down, which is positive. Sites are now loading without issue.
We're currently aware & investigating an issue with the ganaraska cluster. It appears to be a disk IO issue that is causing sporadic outages & slow performance. We have the data centre investigating. Please note that this is only affecting the ganaraska cluster, which is approx. 20 sites.