What It Actually Takes to Network a Live System That's Never Allowed to Go Offline

https://hackernoon.imgix.net/images/Cj9ekT7aeOObvn1smjzED8z4yIw2-6b83srr.png

There is a version of network engineering where you get a maintenance window. Somebody sends a calendar invite, everyone agrees that Sunday 2 a.m. to 6 a.m. belongs to you, and if you overrun by twenty minutes the worst thing that happens is an apologetic email.

I have spent more than a decade doing the other version. The systems I build carry video surveillance, public address, intercom, and emergency two-way communication for a large public transit network. The stations stay open. Trains keep running underneath the work. There is no window where the system is allowed to be down, because "down" means a passenger presses a call button in an elevator and nobody answers.

That single constraint reshapes every technical decision that follows. I want to walk through what it actually changes, because I think a lot of the network design advice floating around assumes a greenfield that most infrastructure...

Copyright of this story solely belongs to hackernoon.com. To see the full text click HERE

Read more