Skip to content
Register

The outage that taught me to stop trusting status pages

General Discussion by Fritz48 4 replies 219 views
Fritz48

Member OP

Datenschutz First

8 #1

Last Tuesday, Contabo showed all green for 6 hours. My server was completely unreachable. As far as I know, their status page (https://status.knownhost.com) checks internal network only, not actual customer connectivity. This is a transparency issue under GDPR Article 5(1)(a) - accuracy principle. I requested their monitoring methodology. They declined. I migrated to KnownHost the same week.

Neuland. Aber schnell.
Emre10

Member

#2

Fritz48 said:
As far as I know, their status page (https://status.knownhost.com) checks internal network only, not actual customer connectivity.
Bro same thing with Hostinger since 2 years. Status page green but server down for me. I tweet them, they say "check your connection bro". My connection fine, I run traceroute from 3 location. They not care. Important lesson: status page for them, not for us.

salavi

Member

Threat Modeler

#3

Threat model: assume compromise of status page itself. Blast radius includes all customers making go/no-go decisions on false data. Attack surface: status page API often on same infrastructure as production. Zero trust means verify independently. I run parallel monitoring against provider API endpoints, not their status page. Zero trust as personality, not product.

kenji3

Member

#4

I believe HostHatch in Singapore had similar issue in 2023. Their status page hosted on CloudFlare, actual infrastructure on separate network. Page stayed green during routing failure to APAC. I prefer providers who use independent status infrastructure, perhaps even different provider. Humble opinion: status page should fail open, not fail closed.

conbini > datacenter snacks
olespete

Member

Trust No One

#5

WARNINGS: status page green while down means THEY CAN EDIT REALITY

What could go wrong:

  • Manual updates "forgotten" during incident
  • Automated checks only ping management interface, not customer VMs
  • Deliberate delay to reduce ticket volume
  • Your monitoring becomes their early warning, free labor

Fail2ban your heart. Firewall your trust. 👀

airgapped, encrypted, faraday'd, still worried

Post a reply

You need an account to reply. Log in or register to join the conversation.

Post reply Preview Save draft