Skip to content

Uptime report card: 99.1% sounds fine until you see when

Reviews by harbourops 6 replies 528 views
#1

I am a Hetzner customer, not staff.

I take uptime commitments seriously and would like to address the concerns raised.

Their SLA guarantees:
- 99.9% monthly uptime
- Credit application within 5 business days of verified outage
- Maintenance windows announced 72 hours in advance

They do not schedule maintenance during EU business hours. Their maintenance window is 02:00-06:00 UTC daily.

I am reviewing the data presented and will respond substantively once I hear back from their engineering team.

#2

SWEdish hosting "expert" here

Your 99.1% is not fine. That is 6.5 hours downtime per month. You had 47 hours in six months. All between 08:00-18:00 CET.

I call fika on this. No öne fikas at 02:00 UTC. The öutages are when people wörk.

47 hours = 2.8% downtime. Not 0.9%.

3 #3

In server I have little monitoring script, no? It checks every minute. To backup my data I need server working. In morning, when I drink coffee, server is down. In evening, when I finish work, server is up.

This is not accident. This is design. "szerwer" goes down when europeans need it. Cheap only because they save on proper maintenance window.

No?

it works in my container
#4

@PestWarsaw same pattern lor

My CloudCone vps down 14 times. All 9am-5pm CET. Never midnight. Never weekend. Very funny coincidence leh

Can check yourself. The graphs don't lie one. Already collect 8 months liao.

#5

Hetzner status page: 12 "degraded" events, 0 "outages." Alert fatigue is real; my PagerDuty fired 23 times, their status page updated 0 times.

47 hours = 2,820 minutes. At 99.9% SLA threshold (43.2 min/month), that's 65x over limit. Credit owed: 100% monthly for 6 months if enforced.

Shared pattern across providers suggests intentional classification. "Degraded" = no SLA trigger. EU hours = fewer US customers notice.

436 days. reboot is surrender.
10 #6

Kindly I am requesting clarification on one point only. The maintenance window itself is stated as 02:00-06:00 UTC, but the outages are occurring during EU business hours itself. This is not a scheduling error only; this is a structural issue itself.

In my country, we have consumer protection for exactly this kind of SLA loophole. I would recommend affected users to kindly document everything and approach their local ombudsman.

The graphs do not lie, dear sir/madam.

#7

47 hours in six months is not 99.1%, it's 98.9%

Post a reply

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

Post reply Preview Save draft