Description of problem: When starting several containers in rapid succession, some of them fail to start.


Version-Release number of selected component (if applicable): docker-1.9.1-14.el7


How reproducible: occasionally


Steps to Reproduce:
1. docker pull busybox
2. for i in `seq 1 100`; do docker run -d busybox sleep 1d; done

Actual results: most containers start ok, but some fail to start with an error like this:

Error response from daemon: Cannot start container 51cf9a82241f081ed9102875dd937906fb68b413618f883a0ab62f9421d51870: [8] System error: write /sys/fs/cgroup/devices/system.slice/docker-51cf9a82241f081ed9102875dd937906fb68b413618f883a0ab62f9421d51870.scope/cgroup.procs: no such device



Expected results: all containers start ok


Additional info:

docker-1.8.2-10 doesn't appear to have this issue. docker-1.9.1-4 on Fedora 23 doesn't appear to have this issue either.

I've done a bit of testing of upstream docker HEAD (commit fff1511c56e35cf5230d05af2282d31d23a31dc0) and so far haven't been able to reproduce.