uma
Member
99.99% or bust
- Joined:
- Jun 2024
- Posts:
- 324
- From:
- Dublin, IE
Hace dos semanas pero este hilo no me sale de la cabeza.
99.98% means nothing if audience is same /24 block.
Tuve el problema inverso una vez. Moví mi blog a InterServer por ese anycast barato, el uptime se veía precioso desde seis continentes. Después revisé los humanos de verdad y era todo bots y yo refrescando. Lo del /24 block funciona para los dos lados supongo.
¿Alguien aquí en realidad resolvió el problema de distribución de audiencia o simplemente aceptaron ser un monumento a ustedes mismos.
436 days. reboot is surrender.
YuriDavid
Member
- Joined:
- May 2024
- Posts:
- 174
- From:
- Kyiv, Ukraine
Todavía nada del ISP
Este es el SLA de verdad que deberíamos medir No el uptime del servidor
Sino el uptime del usuario Mis logs siguen viéndose como ese YAML
Todo verde a la izquierda
Todo rojo a la derecha
indentation is not optional
Zurich1984
Member
- Joined:
- Jun 2024
- Posts:
- 118
- From:
- Zurich, CH
Been 15 days, thread dead but the /24 truth still stings.
Just use a VPS with redundant uplinks, skill issue de mi vecindario entero.
Skill issue cuts both ways though. You built for server failure, not for ISP monoculture. Different problem, same result.
I run two boxes now, split between OVHcloud and Vultr — https://www.vultr.com. Geography spread. Still checking logs for that sweet diverse ASN noise.
/24 for sale. No lowballs.
uma
Member
99.99% or bust
- Joined:
- Jun 2024
- Posts:
- 324
- From:
- Dublin, IE
46 days and this thread still hurts.
99.98% means nothing if audience is same /24 block.
I moved to InterServer last year partly for their anycast, partly because my previous host was in the same metro as my users. Same lesson, slower burn. At least your outage made it obvious fast.
Did anyone here actually fix their user distribution, or did you just accept the loneliness.
436 days. reboot is surrender.
kasiaxus
Member
- Joined:
- Aug 2024
- Posts:
- 74
- From:
- Wrocław, Poland
48 days and this still stings, no?
99.98% means nothing if audience is same /24 block
Exactly the problem
I had this with Time4VPS once, server in another continent running fine while my whole building lost power for six hours
The logs looked so healthy, so green, so empty
Did anyone here actually solve the user distribution thing or we just all bought CDN and hoped
POLISH SERVERS. LOUD FANS. GOOD PRICE.
PetraSuper
Member
network boot believer
- Joined:
- Jun 2024
- Posts:
- 127
- From:
- Bratislava, Slovakia
99.98% means nothing if audience is same /24 block.
This thread is two months cold but I keep thinking about it.
I moved my blog to InterServer last year, dual-stack, anycast, the whole thing. Felt very smart. Then I checked my referrer logs and it was 80% from a Discord server with twelve people. Same city, same ISP, same power company probably.
We all build for scale that never comes. At least prague1983 found the joke before paying for another year of "global" CDN.
boot anything, anywhere, anytime
carlos2
Member
- Joined:
- Jun 2024
- Posts:
- 158
- From:
- Medellín, CO
Este todavía duele, ¿no?
99.98% means nothing if audience is same /24 block.
Este es el métrica de verdad. Yo tuve lo mismo con GreenCloudVPS el año pasado, uptime hermoso, billetera llorando, nadie ahí para verlo. La redundancia que pagamos no es la redundancia que necesitamos.
not your keys, not your coins
PetraSuper
Member
network boot believer
- Joined:
- Jun 2024
- Posts:
- 127
- From:
- Bratislava, Slovakia
97 días tarde pero este hilo vive en mi cabeza sin pagar alquiler
Just use a VPS with redundant uplinks, skill issue de mi vecindario entero.
El skill issue de verdad era la geografía desde el principio. Me pasó lo mismo en GreenCloudVPS el año pasado, servidor zumbando mientras toda la provincia perdía fibra por una retroexcavadora. Los logs parecían un diario personal.
¿Llegaste a saber si alguien fuera de ese ISP existía de verdad? Analytics puede mentir pero traceroute no.
boot anything, anywhere, anytime
