Okay, network mayhem! High system CPU usage within the main router, and this is interfering with the ability of user-mode processes to run. And the most jarring issue, the fact of the matter is that for this particular problem, this looks to be caused… almost entirely internally? Well, the point in hand is that something buggy is going on with IPv6 protocol responses, but there’s not a significant amount of network traffic being generated. At least not externally, like it is all going internal but we don’t know where?
But, good news, just do /etc/init.d/network restart and all is fine.
It took a real long time to run the command with the user-mode CPU
starvation, but after running it, all was well.
Just remember, this command is the best one to run when the network has gone wild. And don’t worry, it is designed to never disconnect the local IP address, so if it fails, you will still be able to login locally for administrative purposes. Just that it will restart all services. Best method is to do this, trying to restart sub-services piecemeal like dnsmasq, odhcpc, dropbear, etc. usually doesn’t go too well because it creates other conflicts.