'Cultlike' Devotion: Apple Once Refused To Join Open Compute Project, So Their Entire Networking Team Quit (businessinsider.com)

mattydread23 writes: Great story about the Open Compute Project from Business Insider's Julie Bort here, including this fun tidbit: "'OCP has a cultlike following,' one person with knowledge of the situation told Business Insider. 'The whole industry, internet companies, vendors, and enterprises are monitoring OCP.' OCP aims to do for computer hardware what the Linux operating system did for software: make it 'open source' so anyone can take the designs for free and modify them, with contract manufacturers standing by to build them. In its six years, OCP has grown into a global entity, with board members from Facebook, Goldman Sachs, Intel, and Microsoft. In fact, there's a well-known story among OCP insiders that demonstrates this cultlike phenom. It involves Apple's networking team. This team was responsible for building a network at Apple that was so reliable, it never goes down. Not rarely -- never. Building a 100% reliable network to meet Apple's exacting standards was no easy task. So, instead of going it alone under Apple's secrecy, the Apple networking team wanted to participate in the revolution, contributing and receiving help. But when the Apple team asked to join OCP, Apple said 'no.' 'The whole team quit the same week,' this person told us."

Comments

Never Down

[speedplane]

This team was responsible for building a network at Apple that was so reliable it would never down. Not rarely — never.

Leave it to business insider to make ludicrous claims about network availability. If Apple's network had 99.99% uptime, and it would cost ten billion dollars to add another 9 to it, I'm pretty sure they'd rather pocket that money than spend it on more redundant switches/routers.

Re:Never Down

[cheater512]

Full redundancy still has outages, even significant ones.

In my experience the more layers of redundancies, the more edge cases you need to catch.

Re:Never Down

[AK Marc]

Full redundancy has piles of single points of failure. I've seen a BGP flap take out a Fortune 100 company with N+1 redundancy that cost billions. Redundancy increases complexity, and there's always a point where the "redundancy enabling" technology becomes a single point of failure.

Re:Never Down

[Gussington]

Um, you do realize that there are networking technologies to protect the network from practically every scenario that you mentioned?

There are no free lunches. You either have a simple network that could fail and is easy to understand and fix, or a complex network that could also fail but is a nightmare to understand and fix. The other big issue with the latter network, is the size and complexity makes upgrades and patching difficult and expensive, and if people leave it's difficult to bring them up to speed on how it works. The results is that it costs more to maintain than you would lose with a few hours downtime each year. So there is no such thing as a network that never goes down, you either have one that is cost effective or you don't.

Re:Never Down

[AK Marc]

And no, redundancy doesn't make things harder as long as it's implemented properly

"properly" by your definition is prohibitively expensive. Almost nobody does it. Realistic redundancy leaves lots of gaps and holes. And in many cases, active/standby is dangerous. HSRP, STP, and many other protocols are active/standby with errors in the standby allowing massive networking failures. And, of course, the protocol to manage that redundancy is a single point of failure. You could abandon HSRP to avoid that single point of failure, and instead have multiple gateways and every endpoint running a dynamic routing protocol but that just moves the single point of failure to whatever routing protocol you pick, and isn't generally done for a variety of very good reasons.

Nope, the simplest network is often more reliable than the rube goldberg redundant networks I've seen experts like yourself put together. KISS is one of the first rules, and the more you know, the more it matters. KISS. Anything else is expense for the sake of complexity.

Re:Reasonable

[DogDude]

We use Chinese made hardware because it works better than US made stuff (Cisco). If the Chinese gov't wants to spy on our business, they can have at it. It's worth it to us to have reliable equipment.

Re:Reasonable

[93 Escort Wagon]

Cisco buys from the Chinese. All the best stuff is made in China

All the worst stuff too.