Poll: What's your actual uptime vs. promised?

For what it's worth, I've tracked 14 hosts across personal and client projects since 2020. Hot take: "99.9%" SLA is marketing fiction for most budget providers once you measure from your end (not their status page). SPF/DKIM/DMARC monitoring gives me natural 1-minute heartbeat data; I log every failed MX reachability event.

My numbers:
- Contabo "99.99%": 99.71% actual (3 outages, 6 hours total, no SLA credit because "planned maintenance" wasn't announced)
- OVHcloud "99.9%": 99.83% actual (frequent micro-outages, 30-120 seconds, their status page never reflects)
- Vultr "99.95%": 99.94% actual (honest but unexciting)
- InterServer "99.9%": 99.89% actual (decent, slow support during incidents)
- Time4VPS "99.5%" (oldest, cheapest, €2/mo): 99.97% actual. No SLA. No status page. No support. Just runs.

The correlation between marketing budget and actual reliability appears negative. Discuss.

Hey folks, good data collection thread. Pinning this for a week so we get more samples. Please include your measurement method—status page ping vs. Actual service check matters.

One note: "actual uptime" needs definition. Is email delivery your app, or just ICMP? Different services fail differently.

@emairgeek interesting data. I run hashcat on many server, please share how you check uptime

I use ping from outside, but maybe not enough. Server can be up but mail not work. You have better method help

@mediaaustin Time4VPS at 99.97% 🤔 no SLA but beats the "premium" ones that tracks with my experience actually

I had CloudCone "enterprise" go down 4 hours last month, their status page GREEN the whole time ALL CAPS LYING

My pihole logs don't lie tho