Live data from Hacker News

Cloudflare had a partial outage

cloudflare.com

201–210 of 448 posts

Re: Cloudflare had a partial outage

#202
post #138

Earlier quoted context omitted.

We are not using cloud flare. But our domain is also not accessible. We are using digital ocean's DNS service for propagating our IP. Does the DigitalOcean's DNS service depend on Cloudflare service?

Yes https://www.cloudflare.com/case-studies/digitalocean/

That link isn't accessible from where I am right now.

Alanis Morissette agrees that this is ironic.

Re: Cloudflare had a partial outage

#203
post #24

This should hopefully drive home the idea of why HN shouldn't be cheering on Cloudflare's slow takeover of the internet.

CLoudflare just offer great services. Its straight up fact that even their free model is extremely generous. There is no big conspiracy to 'take over the internet' but when the product is good, the product is good.

Not a 'big conspiracy'. It's the business model, isn't it? Or isn't CF going for the biggest marketshare and maximizing profits on that, like all the others?

Re: Cloudflare had a partial outage

#204
post #102

At times like this and the big Fastly outage roughly a year ago, choosing to host on a simple, independent bare-metal box doesn't seem like such a bad strategy (as long as one has backups for disaster recovery, of course). Sure, other things can cause downtime in that kind of infrastructure, but at least my service isn't likely to be taken offline by someone else's configuration error or deployment gone wrong.

I'm on Linode. Linode is down because Cloudflare is down. Can't login to their control panel, etc. You'd need to go fully independent and roll your own, with zero dependencies, to really make this work.

Yeah, they'd also be dependent on their ISP still if they're "fully independent". Good luck dealing with massive traffic spikes on a single bare-metal box and good luck maintaining a similar uptime to cloudflare's 98.84% uptime lol

Re: Cloudflare had a partial outage

#205
post #119

Earlier quoted context omitted.

What is your setup to where you are isolated from "other person making a mistake"? Even if you're a box in a colocated datacenter you're still able to get knocked off the net from some maintenance on the surrounding pipes. Hell, hosting your own box doesn't cause Comcast DNS issues to not knock off a bunch of people either. I do think there is some holistic overview of hosting stuff on the internet, where you could l…

Fair point. Still, based on my anecdotal experience using leased dedicated servers, mistakes at that networking layer seem to happen less often than mistakes that take AWS us-east-1 or one of the big CDNs offline.

It does feel like more "hosted" environments are trying to do more fancy stuff inside the network, so have more failure cases. Or perhaps services that do a lot of things, even if you end up just using simple server components.

I still have a fun memory of half of IBM Cloud's servers falling over, meaning that our production app was luckily still up but our staging server fell over. I could get to their website, but their login stuff was all messed up. I believe that one was also a "routing stuff got messed up" issue....

Re: Cloudflare had a partial outage

#206

At times like this and the big Fastly outage roughly a year ago, choosing to host on a simple, independent bare-metal box doesn't seem like such a bad strategy (as long as one has backups for disaster recovery, of course). Sure, other things can cause downtime in that kind of infrastructure, but at least my service isn't likely to be taken offline by someone else's configuration error or deployment gone wrong.

> at least my service isn't likely to be taken offline by someone else's configuration error or deployment gone wrong.

It's likely to be taken offline by yourself more often than not though.

Re: Cloudflare had a partial outage

#208
post #142

At times like this and the big Fastly outage roughly a year ago, choosing to host on a simple, independent bare-metal box doesn't seem like such a bad strategy (as long as one has backups for disaster recovery, of course). Sure, other things can cause downtime in that kind of infrastructure, but at least my service isn't likely to be taken offline by someone else's configuration error or deployment gone wrong.

My domains with DNSimple, and my servers are on Hetzner. There should be no dependency on CloudFlare, and yet they are down too.

Good to see how decentralised the Internet truly is.
Post reply on HN