kaeros.app
Operational 344 ms 95.00% 90 days ago today
Status
Is anything broken?
All systems normal
Services
What we watch
Each bar is one day. Green is a day with no failed check; amber means it was up but slow; red means at least one per cent of checks failed. A gap means we were not watching yet.
erioun.com
Operational 53 ms 99.95% 90 days ago today
jobopportunitiesapi.org
Operational 134 ms 98.70% 90 days ago today
tzekos.eu
Operational 53 ms 99.99% 90 days ago today
Clients
Operational 35 ms 100.00% 90 days ago today
Internal
Operational 121 ms 99.69% 90 days ago today
Infrastructure
The machines underneath
Three servers, all ours, all in the European Union. Two are rented in Warsaw and carry everything a client touches. The third we own and run ourselves in Thessaloniki, and it holds internal tooling only. The company site and its support system deliberately do not live on the same machine as the products — one outage should not take down the thing you would use to find out about it.
Application host
OperationalProducts and client workloads
- RegionEU · Poland
- Operated byTZEKOS.EU
- BackupsNightly, encrypted, off-site
- Last health report4 min ago
Studio host
OperationalThis site, its portal and image builds
- RegionEU · Poland
- Operated byTZEKOS.EU
- BackupsNightly, encrypted, off-site
- Last health report2 min ago
Internal host
OperationalStudio tooling, AI and GPU workloads — nothing a client depends on
- RegionEU · Greece
- Operated byTZEKOS.EU
- BackupsNightly, encrypted, off-site
- Last health report41s ago
Exact processor, memory and disk figures are not published here. They tell an attacker how much headroom there is to exhaust, and they tell a customer nothing they cannot read off the uptime history above.
Updates
What happened, and what we did
Written by a person, never generated. Incidents stay here after they are resolved — a status page that deletes its own history is a marketing page.
-
Backup & Restore testing
Fri 7 Aug - 21:00 EET: Backup and restore testing on two machines. Some delays and outages could be expected.
-
Scheduled CPU stresstest
We are conducting a stress test to identify failure points inside our app by overloading the CPU. That test will run for around 3 hours between 5:00 EEST and 8:00 EEST.
What we promise when it breaks
A response within 24 hours, written into the contract. Deliberately not an uptime percentage — a response is something one person can guarantee, and it is what actually matters at eight in the morning when the booking form is down.