Starting a network interface issues a mountnfs regardless if the network interface has any associated nfs mounts

Binary package hint: initscripts

Informations:
System:
Server with 2 NICs and Dapper installed:
Linux rad3 2.6.15-27-server #1 SMP Sat Sep 16 02:57:21 UTC 2006 i686 GNU/Linux

initscripts:
  Installed: 2.86.ds1-6ubuntu32

Network Configuration (/etc/network/interfaces):
auto bond0 vlan169 vlan777

# Server VLAN
iface vlan169 inet static
   vlan-raw-device bond0
   address 172.x.y.z
   netmask 255.255.255.224
   gateway 172.x.y.z

# NFS VLAN
iface vlan777 inet static
   vlan-raw-device bond0
   address 192.168.c.d
   netmask 255.255.255.0

# Administration VLAN
iface bond0 inet static
   up ifenslave bond0 eth0
   up ifenslave bond0 eth1
   post-down modprobe -r bond0
   address 192.168.a.b
   netmask 255.255.254.0
   broadcast 192.168.a.b

NFS Mount (/etc/fstab):
r150-nfs:/vol/radiusbackup /backup nfs rw,bg,vers=3,tcp,timeo=600,rsize=32768,wsize=32768,soft,intr 0 0

Problemdescription:
If the Server boots, the mountnfs script is issued as soon as bond0 is up and hangs the bootprocess cause r150-nfs is not available at that moment. We have tracked back this behaviour and it’s caused by the “ifup –a”-command, which starts the mountnfs script. mountnfs now tries to mount all nfs mounts listed in /etc/fstab. Since the NFS-Server is located in our NFS-VLAN, and that network relies on the interface “bond0” the mount fails and hangs the further boot process.

Workaround:
We have worked around the problem by adding
[ "$IFACE" == "vlan777" ] || exit 0
at the beginning of mountnfs but that’s a really bad workaround…