|
|
Post by stanleyb on Jun 8, 2021 19:10:06 GMT
I have always had my doubts about relying on Cloud storage instead of having storing stuff on a hard drive at home or work. I lost most of today as far as work is concerned. My internet service provider also went down with the error 503 code and didn't come back to life till around 5:40PM
|
|
|
|
Post by MartinT on Jun 8, 2021 19:17:20 GMT
|
|
|
|
Post by stanleyb on Jun 8, 2021 19:22:13 GMT
And all that begs the question whether Big Tech have sold us a vulnerable system as a robust solution.
|
|
|
|
Post by MartinT on Jun 8, 2021 19:28:12 GMT
These are vast and complex cloud systems. All the major vendors like Azure, AWS, Google have had outages. The fact that few people have heard of Fastly suggests that they've been pretty reliable to date.
Going cloud is always a risk, but you're buying not just the service itself but their ability to get back online rapidly, and that's where the best services are strong. Imagine the time it could take to recover a crashed standalone web server with a few TB of content? Could be several days.
|
|
|
|
Post by stanleyb on Jun 8, 2021 19:40:17 GMT
The outage raises questions about the consolidation of internet infrastructure. The outage, like the pandemic, should us of our vulnerabilities and the need to design with not just the slickest outcomes but also the worst-case scenarios in mind.
|
|
|
|
Post by MartinT on Jun 8, 2021 19:46:27 GMT
Yes it does. You almost wonder whether the biggest sites will hedge their bets by straddling two vendors in future, so that they can retain some functionality if one goes down.
I think one of the biggest single points of failure would be the root DNS servers. Take those down and there would be a gradual shutdown of absolutely everything as DNS records expire and need renewal.
|
|
|
|
Post by MikeMusic on Jun 8, 2021 20:06:47 GMT
Once upon a time I wasn't paranoid
This doesn't bode well for the future
|
|
|
|
Post by Slinger on Jun 8, 2021 20:12:26 GMT
Cloud computing provider Fastly, which underpins a lot of websites, said it was behind the problems. The firm said there had been issues with its global content delivery network (CDN) which it was fixing. In a statement, it said: " We identified a service configuration that triggered disruption across our POPs (points of presence) globally and have disabled that configuration." That " service configuration," sounds a bit like a deliberate act to me. Here's the bit I really don't understand though. Fastly runs what is known as an " edge cloud", which is designed to speed up loading times for websites, as well as protect them from denial-of-service attacks and help them when traffic is peaking. Now I'm sure that's oversimplified, but why would you have an ancillary system that can't be bypassed? i.e. It fails, so you switch it out and go around it, and things run a bit slower until it's fixed as opposed to a complete outage. You may remember this though, from yesterday... You will all be very proud of me I managed to take a photo with my phone, attach it to a message, only took 3 goes - and save as a draft - as I couldn't work out how to send but picked it up anyway on the laptop to send from there ! I am a genius !  Oh well. That's the internet f*cked then.  |
|
|
|
Post by MartinT on Jun 8, 2021 21:01:55 GMT
Similarly, ProBoards run with the help of Cloudflare. There seems to be a lot of reliance on these intermediaries.
|
|
|
|
Post by stanleyb on Jun 8, 2021 21:05:47 GMT
Hmmm... Can we start a conspiracy theory that Mike inadvertently caused the outage?
|
|
|
|
Post by MartinT on Jun 8, 2021 22:12:48 GMT
I believe you just have!
|
|
|
|
Post by MikeMusic on Jun 9, 2021 10:24:29 GMT
Yus guvnor it's a fair cop
Throw on the bracelets
|
|
