[CentOS] centos 7, docker, NFS and uid = -2

      I created a centos 7 docker container in which I want to mount a
NFS share in. Said share is owned by user virtual with uid 1200. So I
do some exporting (docker container is in 172.17.0.0/16):

spindizzy> cat /etc/exports
/export 10.0.0.0/24(ro,fsid=0,no_subtree_check,sync)
172.17.0.0/16(ro,fsid=0,no_subtree_check,sync)
[...]
/export/mail 172.17.0.0/16(rw,root_squash,no_subtree_check,sync)
spindizzy>

and then some mounting:

bash-4.2# id virtual
uid=1200(virtual) gid=1200(virtual) groups=1200(virtual)
bash-4.2#  nfsstat -m
bash-4.2# mount.nfs4 -v -o proto=tcp fileserver.in.kushana.com:mail
/var/spool/vmail
mount.nfs4: timeout set for Wed Oct  8 03:11:43 2014
mount.nfs4: trying text-based options
'proto=tcp,addr=10.0.0.18,clientaddr=172.17.0.29'
bash-4.2# nfsstat -m
/var/spool/vmail from fileserver.in.kushana.com://mail
 Flags: rw,relatime,vers=4.0,rsize=131072,wsize=131072,namlen=255,hard,proto=tcp,timeo=600,retrans=2,sec=sys,clientaddr=172.17.0.29,local_lock=none,addr=10.0.0.18

bash-4.2# ls -l /var/spool/
total 12
drwxr-xr-x 2 root       root       4096 Jun 10 01:11 lpd
drwxrwxr-x 2 root       mail       4096 Oct  8 02:15 mail
drwxr-xr-x 2 4294967294 4294967294 4096 Oct  3 17:34 vmail
bash-4.2#

And am getting the famous share being mounted as nobody issue. In the
client (docker container), I went in /etc/idmapd.conf and set the
Domain to be the domain where the NFS server is at (in.kushana.com).
Now, how do I restart idmapd? It is a systemd service
(/usr/lib/systemd/system/rpcidmapd.service, which uses
/etc/sysconfig/nfs).