[aoetalks]

Really curious to see how a single (albeit large) DC issue caused a global outage.

Further; was the DC not protected by redundant power?

Also even more interested to see how this arc flash even happened. Presumably these were experienced engineers working with obviously dangerous power levels.

[jeffbee]

> was the DC not protected by redundant power?

If this topic is of great interest to you, US Patent 10,678,647 could be fun.

[shadowgovt]

> how a single (albeit large)

That's the reason. In principle, Google distributes execution and data storage for its mission critical applications. And most of the time, they do a pretty good job of that.

But (especially with the hardware crunch in recent years brought about by various effects, including COVID and Bitcoin mining putting pressure on the market) their ability to build out new datacenter capacity has stalled, and as a result, things sometimes end up bunched up in suboptimal ways, and a high capacity data center can become a single point of failure even though it shouldn't be.

> Was the DC not protected by redundant power?

Depending on where the flash occured, it may have been at the point the redundancies feed into the building.