You are not logged in.
Hey folks, I’m dealing with a pretty weird problem at the moment. It seems the Bay Area, California (where I live) is having a pretty massive unplanned internet outage. My laptop is connected to my local wifi router on my LAN however the WAN is down. I noticed that my CPU monitoring widgets were reporting a ton of CPU usage and htop revealed that systemd-resolved was basically consuming 100% of my CPU and blocking most all other processes from completing. I couldn’t even get my computer to poweroff without using the sysrq REISUB.
Has anyone noticed this before? Is this something I should report upstream to systemd? Or is there any other troubleshooting I should do while I have the time?
I suspect this should be rather easy to reproduce, however will be annoying lol. Maybe in a VM it’ll be easier.
Offline
I assume the cause is that something (networkmanager) tries to resolve a WAN domain to indicate the online status and I'd be less than surprised if resolved threw a hissy fit because of the ingenious idea to read a file through a socket…
local wifi router on my LAN
that most likely will provide you w/ cached DNS responses anyway (you can check the response time w/ "dig @router_ip_here google.com", if it's around 0ms (at least on subsequent queries and once the WAN is up again…) systemd-resolved is about as useless as tits on a boar.
Offline