There is a very important concept for Cloud Services, which is called “High Availability”, and basically the way you design your Infrastructure is such that the Service being provided while is tolerant to failure, it minimizes the downtime.
Although a service can go off, its ok, one will pop up. This often aims for 99.9% availability, although there are others that models in which maximum downtime is about
This is in contrast to for example Fault Tolerant which are systems in which they cannot afford to fail, like Air control towers or hearth rate monitors, etc.
So my question for Blizzard, is why has a High Availability Architecture not been implemented for WOW ?, You could easily “patch” on a server, have a mirror server, then the the patching as seamingly easy as just having to relog, and having an extra server hanghing just in case.
I know this costs money, you need good Cloud Architects and Engineers and the infrastructure, but for real, your standard has to change regarding maintenance.
Why is it still ok at 2023 to have this long maintenance and downtime ? No one is asking for 99.9999999% (nine nines) which are basically milliseconds per year. But a good 99.9% would be good to aim.
That would mean 10 minute downtime per week. And i think im being extremely generous here.
Why are Blizzard’s standards, still 20 years behind when it comes to infrastructure ? This is a serious question.
The implementation of such a high available system is one yaml document away and blizzard’s good will. The technology is there since several years ago
P.S. I hope this post doesnt get banned for “Innapropriate”