---
title: CEPH slowness / failure, dead physical host
date: 2020-06-23 22:00:00
resolved: true
resolvedWhen: 2020-06-24 07:00:00
severity: down
affected:
  - Network
  - Storage Cluster(s)
  - Virtual Machines (DatacenterLight, IPv6OnlyHosting)

section: issue
---

Network failure in internal network gradually degraded the CEPH (storage)
cluster, which ended up making VMs unresponsives. One physical host hosting VMs
went down during the night. Service gradually went back between 4 AM and 7 AM.

More details to come.