I know that this is a rather old question; however, this occured when I updated from CentOS - Extras docker to Docker's own docker-ce rpm. The service file located at /usr/lib/systemd/system/docker.service had the following lines hardcoded into the ExecStart line:
--add-runtime docker-runc=/usr/libexec/docker/docker-runc-current \
--default-runtime=docker-runc \
--exec-opt native.cgroupdriver=systemd \
--userland-proxy-path=/usr/libexec/docker/docker-proxy-current \
By default, the docker-ce package does not come with the same set of lines and the file paths are different. After upgrading I added the following and the containers respected my start commands with no data lost as expected:
--add-runtime docker-runc=/usr/bin/docker-runc \
--default-runtime=docker-runc \
--exec-opt native.cgroupdriver=systemd \
--userland-proxy-path=/usr/bin/docker-proxy
Hope this helps someone else!