I was changing some network config on our hosts and ran in to ifupdown-ing a bridge and than all the tap devices are disabled ( http://forum.proxmox.com/threads/13...king-retart-Now-Guests-have-no-network-access ).
So I build a quick and dirty fix for it:
And added a post-up hook to my /etc/network/interfaces:
But now I'm also connecting tap devices to vmbr0v1 post-up to vmbr0, but vmbr0v1 never went down in the first place.
So is it possible to add the vmbrXvY bridges explicitly to my /etc/network/interfaces, that way I can give them their own post-up directive
Should it look like this?:
(offcourse I have to edit the script to support these new type of bridges)
So I build a quick and dirty fix for it:
Code:
#!/usr/bin/env bash
hostname=`hostname`
script_name=`basename $0`
bridge_number=${script_name: -1:1}
result=`grep vmbr$bridge_number /etc/pve/nodes/$hostname/qemu-server/*.conf \
| sed -E "s/^.*server\/([0-9]+).conf\:net([0-9]+)\:\s.*bridge\=(vmbr[0-9]+)(,tag=([0-9]+))?$/brctl addif \3v\5 tap\1i\2/" \
| sed -E "s/(^.*vmbr[0-9])v(\s.*$)/\1\2/"`
while read -r line
do
# get status for vm
vm_id=`echo $line | sed -E "s/^brctl addif vmbr[0-9]+(v[0-9]+)? tap([0-9]+)i.*$/\2/"`
echo "Checking if $vm_id is running."
status=`qm status $vm_id`
# check if vm is running
if [[ $status == "status: running" ]]; then
# remove trailing v when there is no tag
if [[ ${line: -1:1} == "v" ]]; then
line=${line: 0:-1}
fi
echo "$line"
$line
fi
done <<< "$result"
# always exit with succes otherwise the device will be completely down
exit 0And added a post-up hook to my /etc/network/interfaces:
Code:
auto loiface lo inet loopback
auto eth0
iface eth0 inet manual
auto vmbr0
iface vmbr0 inet manual
bridge_ports eth0
bridge_stp off
bridge_fd 0
post-up /usr/local/bin/post-up-vmbr0
auto eth1
iface eth1 inet static
address 172.18.100.1
netmask 255.255.255.0
auto vmbr1
iface vmbr1 inet static
address 172.18.100.1
netmask 255.255.255.0
gateway 172.18.100.11
bridge_ports eth1
bridge_stp off
bridge_fd 0
post-up /usr/local/bin/post-up-vmbr1But now I'm also connecting tap devices to vmbr0v1 post-up to vmbr0, but vmbr0v1 never went down in the first place.
So is it possible to add the vmbrXvY bridges explicitly to my /etc/network/interfaces, that way I can give them their own post-up directive
Should it look like this?:
Code:
[COLOR=#333333]iface vmbr0v4 inet manual[/COLOR]
[COLOR=#333333] bridge_ports eth0.4[/COLOR]
[COLOR=#333333] bridge_stp off[/COLOR]
[COLOR=#333333] bridge_fd 0
[/COLOR] post-up /usr/local/bin/post-up-vmbr0v4(offcourse I have to edit the script to support these new type of bridges)