Member OP
The README says 99.99% uptime. The SLO doc claims 150Gbps mitigation. The runbook describes a failover that has never been tested. These are aspirational narratives. 340Gbps attack last month, scrubbing center in Frankfurt actually handled 180Gbps before anycast routing loop. Docs said 200Gbps per PoP. Close enough for fiction.
Member
This is a fun little project actually, I tried writing "honest docs" for my NAT VPS setup
# what actually happens
iptables -t nat -A PREROUTING -j LOG --log-prefix "lol good luck"
# documented behavior
"seamless port forwarding"It actually works though
The Windows Event Collector documentation is remarkably accurate about what the service is supposed to do. The "supposed to" is doing quite a bit of work there. I have seen third-party guides that describe registry keys that do not exist in any shipping build.
Documentation is propaganda for cloud dependency, real operators read source code or run their own monitoring, I selfhost BookStack on a repurposed thin client and document nothing because privacy means no paper trail, calling it laziness is corporate Stockholm syndrome
I write doc already for my server-server, but the doc is from 2021 lah, the config already change, I not update, so the doc is fan fiction already, yes, my colleague read and ask why not work, I say go read source code lah
The documentation is server good or server bad? Is depend, the AWS has doc very aspirational, the tutorial is from 2019, the console already change, ¿why not¡ update, how much cost the time for read doc obsolete