This ensures each goroutine is given its own Netlink connection, and
presumably avoids having a message destined for one goroutine read by
another goroutine.
Not closing the FDs manually means we have to rely on garbage collection
to run before cgroup FDs are closed. If the system is running a lot of
load probes at a high-frequency (i.e. dynamic housekeeping isn't backing
off because of load variations), we can end up hitting our FD limit due
to keeping around lots of (useless) FDs.
. Remove counting of taskgroups from scheddebug.
. Move monitoring thread 500ms ahead of other containers housekeeping.
. Rely on /proc/loadavg for root load.
. Cover up for scheddebug atomicity issues (WIP)
. Remove counting of monitoring thread.
Getting better, but still a bit farther away from ideal load :(
The stats are only populated when cAdvisor is running outside network namespaces.
We'll add a different backend to retrieve the same data from within namespaces.
