Google Cloud reported that a cooling-related failure at one of its London buildings began at 10:13 a.m. PT (that's 6:13 p.m. BST). The building hosts a portion of capacity for Google Cloud's europe-west2-a zone.
As of 3:30 p.m. PT, the issue was only partially resolved, with most customers able to launch virtual machines in all zones of europe-west2. Some customers in europe-west2-a zone were still seeing problems with Google Compute Engine (GCE), Persistent Disk, and Autoscaling.
Meanwhile, Oracle said that a subset of cooling infrastructure within its UK South (London) data center experienced issues on Tuesday "as a result of unseasonal temperatures in the region." Some customers were unable to access or use Oracle Cloud Infrastructure resources hosted in the region, including object storage, compute, and block volumes.
As of 10:16 p.m. UTC (3:16 p.m. PT), a subset of the impacted cooling infrastructure had been restored to an operational state. "We're continuing repair work to further reduce operating temperatures and mitigate service impact," Oracle said. "Some services are beginning to observe recovery in health metrics as mitigation efforts continue."