Apr 19, 2026
12:11–13:05 UTC
Partial outage Network · CH-ZRH
DDoS event absorbed — peak 22 Gbps, scrubbed to edge. Fully resolved; follow-up actions tracked internally.
Live infrastructure health across 16 components in four datacenters. Polled every 20 seconds by external probes outside our ASN; measurements are published raw, including the bad ones.
Web control panel · Hypervisor firmware refresh across a single availability zone — instances live-migrated, no reboots expected.
Each bar below is one day, coloured by worst observed state. Hover a bar for the date & per-day summary.
Last 14 resolved events across all components — auto-refreshed on every page load from our probes.
DDoS event absorbed — peak 22 Gbps, scrubbed to edge. Fully resolved; follow-up actions tracked internally.
Asymmetric routing detected at BIX Bucharest. Fully resolved; follow-up actions tracked internally.
Erasure-coding reconstruction throttled after a double disk failure. Fully resolved; follow-up actions tracked internally.
NVMe device failure, RAID rebuild in progress. Fully resolved; follow-up actions tracked internally.
Rebalance of a hot cell into a cooler shelf. Fully resolved; follow-up actions tracked internally.
Hypervisor-level memory pressure investigation — ~11% of hosts affected. Fully resolved; follow-up actions tracked internally.
Object-store index promotion briefly held the write lock. Fully resolved; follow-up actions tracked internally.
Slow-query regression on the metadata service. Fully resolved; follow-up actions tracked internally.
IOMMU group re-mapping required a short guest stun — ~5% of hosts affected. Fully resolved; follow-up actions tracked internally.
Dashboard rendering degraded for a subset of accounts. Fully resolved; follow-up actions tracked internally.
Cgroup-v2 migration revealed a throttling edge-case — ~12% of hosts affected. Fully resolved; follow-up actions tracked internally.
RPKI-ROA validation misconfiguration on prefix via Lumen (AS3356). Fully resolved; follow-up actions tracked internally.
Aggregate service uptime across all 17 components, by month. Months below 99.99 % trigger an SLA credit — see SLA.