By Richard Speed 24 Jun 2019 at 13:07
3 mins ago..
But it's actually 1 hour 3 mins ago.
Dear El'Reg what's going on!
US network services provider Cloudflare has been celebrating its impending tenth birthday with a good, old-fashioned TITSUP*, er, knees-up. Festivities began at around 1100 UTC this morning with website owners around the world noting that their sites had mysteriously become inaccessible for some users. The issue, according to …
3 mins ago..
But it's actually 1 hour 3 mins ago.
Dear El'Reg what's going on!
The "minutes ago" is js-based, and uses your local timezone to perform the computation.
All non-relative times on the website are in UTC - which as it so happens is one hour behind BST, the current time zone in Europe/London, as daylight savings are in effect.
joeW “Same description could be applied to the Internet in general.”
Run your service from a rack-mounted PC with a second back-up machine with a UPS, and duplicate in multiple locations. That way when there is a fault in one location it doesn't impact the entire WEB.
I thought overall network attacks were down. Interesting...
With grreat power comes great responsibility, and the potential to do this by getting a netmask wrong while configuring a route advertisment or filter. It's one of those 'oops' moments that occasionally occurs, and used to result in buying beers at LINX/NANOG meetings. Or just live in network engineer's memories. Like AS7007. Why have all the helldesk phones started ringing? Where is all my traffic going? Fun times.
It happens. It happens more often when people like vendor X's routers because it's config included a roll-back option. We used to edit configs in vi, validate them, then wave the rubber chicken and committ.
They did note on a HackerNews post ( https://news.ycombinator.com/item?id=20262214 ) that Level3 appeared to be the source of a route leak, although it's been edited since.
Screenshot here - sorry if it's competing with ElReg but I can't find any other decent source with a screengrab:
https://www.bleepingcomputer.com/news/technology/cloudflare-is-having-an-outage-affecting-sites-everywhere/
The edit suggests they don't think it's *just* L3 however, so pinch of salt and all that.
Apparently fixed as of 30 mins, ago, which is reflected in two of my dozen sites I look after no longer intermittently alarming at me.
Steven "it always seems to be Level3, doesn't it" R