Skip to content

Just watched my provider's status page lie in real time

General Discussion by lookuppierre 15 replies 4.6K views
8 #1

I am watch the status page for three hour. Three! And always it is say « All Systems Operational ». Meanwhile my site is die, my users are scream, and the provider is pretend nothing is happen.

I take the Screenshot, I save it. Now I have the Proof. How these people can sleep? I am not understand.

Oui, I know, some outage is normal. But to lie about it? That is the different thing. The Honesty is free, no?

Anyone else have this experience with the RackNerd https://www.racknerd.com or others? I am curious if this is the common disease.

prix fixe infrastructure: €5/mo
#2

Xixixi same here lah. I am use the Vultr last month, they are say "all green" but my database is sleep very deep. I am check the status page, still green. I am buy yesterday the new VPS from OVHcloud, hopepully better lah.

The status page is just pancy page, not real I think. Very pancy. Xixixi

#3

Probably 512MB shared on a cPanel box from 2014.

I run my own status page on a dedicated 64GB EPYC box with NVMe. Updates in 12ms. Not bragging, just saying the infrastructure matters when you have 200+ users checking every 30 seconds.

Budget providers cheap out on the status page because it's not "customer facing" until it is.

#4
HeadRich said:
Updates in 12ms

Anyway I wrote a thing at 3am that scrapes provider status pages and diffs them against actual icmp

#!/bin/bash
# status_lie_detector.sh
# idk why i do this
curl -s status.racknerd.com | grep "operational" && \
  ping -c1 racknerd.com || \
  echo "$(date): LIE DETECTED" >> lies.log

Runs on a cronjob from my toaster. Results are predictable.

builds at 3AM, sleeps at noon
2 #5

Cara, this is going to make me crazy. I use Time4VPS and same thing, always "operational" but my site is down. I changed provider, did it this week.

This status page is just for show, that is the true. I moved to GreenCloudVPS maybe — https://greencloudvps.com — my friend say they are honest about problems. I see.

chill infrastructure for chill people 🦫
#6

Status pages are theater, been saying this since 2019.

#7

12ms updates? On a dedicated box for a status page?

#8
Buenos said:
12ms updates? On a dedicated box for a status page?

The box does other things. Status page is one service among several, not the sole workload. 12ms is round-trip from my monitoring node to the status endpoint, not render time.

Point being: if you care about accuracy you don't share the status page infrastructure with the thing it's monitoring. Separate network path, separate power, separate everything. Otherwise you get the green light of death where the page stays up because the failure hasn't reached it yet.

RackNerd probably runs theirs on the same cluster that hosts customer sites. Seen it before.

#9

20 days late but this thread is too familiar

HeadRich said:
Budget providers cheap out on the status page because it's not "customer facing" until it is

This is exactly the thing though — it IS customer facing the second something breaks

I moved from RackNerd to KnownHost last year and the difference was just that KnownHost posted a yellow banner within maybe four minutes of a routing hiccup. Four minutes! Still annoying, but I could stop refreshing and go make coffee instead of guessing

The status page isn't about being pretty, it's about reducing the support tickets from people who just want to know if they should wait or panic

chill infrastructure for chill people 🦫
#10

I have seen this with KnownHost once. The page stayed green for six hours, then suddenly showed a "resolved" notice with a backdated timestamp. Very strange.

Post a reply

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

Post reply Preview Save draft