On Sep 14, 2022, a cabinet in the Singapore data center experienced a power issue that temporarily prevented ~85% of the Linodes from reaching the network. At 18:02 UTC, a power event occurred that impacted both PDUs in a rack. By 18:16 UTC, non-redundant power was restored and the impacted Linodes started to regain network access.
From that point on, our team performed ongoing maintenance to completely replace the impacted PDUs to ensure this did not happen again, and that full redundancy was restored at 01:00 UTC on Sep 15, 2022.
We are actively investigating the source of the power issue and will be discussing this specific incident with the PDU vendors to fully determine the root cause and ensure this will not occur again.