Skip to content

Comment on Proton Global Outage

Comments

Why does a cooling failure in one datacenter brick the entire ecosystem? Reliability 101 to have redundancy here I'd think...

I have to admit, this one has pissed me right off. I build systems that survive datacentre failures all the fucking time, it's inexcusable that a single site going dark removes access to something I pay like 200 euro a year for.

guess it's the same ol' reliability doesn't get prioritized until something like this happens story

Back online for me ~2 hours later. Jesus.

Just for anecdata: I'm using simplelogin to alias my emails and I can see there is a 20ish minute window between when Simplelogin receives an email, to when it finally shows up in my Proton inbox. Normally it's an inconsequential delay.

Not sure if that indicates how much of a backlog of emails Proton has because of this outage, or if the DR infra is just that much slower?

We are currently in the process of shifting traffic to our backups sites

Manually?!

I guess multiple datacenter hosting and automatic switch over should be the resulting solution of this incident

click the words "global outage" on the main status page

AboutSource Built by g1lg1l

Hackerly is an independent reader for Hacker News, built on the public HN API. Not affiliated with Y Combinator.