3 ms·

I'm trying to decide if I am surprised or not that something like this relies on a single central point of failure. (well I assume it's a single point of failur

I'm trying to decide if I am surprised or not that something like this relies on a single central point of failure. (well I assume it's a single point of failure given how this appears to have happened nation wide all at once) You would think that a nation wide cell service would be more distributed and that it wouldn't all go down at once like this. But maybe there are some good enough reasons that there needs to be some central components or systems that make it all work.

tyre 3y ago
Not necessarily a single point of failure. There could be standby servers that were promoted to primary when the primary failed, then fell over with the huge spike in load. Or they were load balanced, multiple failed, and then the others tipped over. In those cases, there were mitigations, but the failures cascaded.