I am running CentOS VMs in Virtualbox using Vagrant on a Windows PC. I've done this for 2 years without issue (initially CentOS 6 then CentOS 7), but in the last few months, I notice my VMs to start locking up frequently. And outputs these types of messages when it gets going again
Message from syslogd@node1 at Jan 25 17:51:55 ...
kernel:NMI watchdog: BUG: soft lockup - CPU#1 stuck for 23s! [C1 CompilerThre:7243]
repeated a number of times.
I think what is happening is the NMI watchdog is waiting for an interrupt, and has a set time to receive, but if a process is holding things up and it doesn't get this interrupt it assumes something has gone wrong and initials a kernel panic, locking things up for a bit.
I've come across options for increasing the kernel.watchdog_thresh=30
in the /etc/sysctl.conf
or disabling the NMI watchdog altogether.
Should I consider trying one of these?
Is the watchdog too eager, or might I have some underlying issue? there's not much on my VMs when this happens.