Virtual router not handing out IPs to my instances

[Carlos Reategui <cr...@gmail.com>]

My virtual router has connectivity problems.

I am able to start an instance but it is not able to get an IP address from
the vr.  If I assign a static IP address (telnet to the 590x port on the
host) to the instance I am able to connect to it.

If I connect to the vr using the link local ip on the host I can see that
it is not able to connect out at all.  I am not able to ping any addresses
either in the same subnet or outside.  I have verified the vif and it is
using the correct one interface:

root@r-4-VM:~# ifconfig eth0
eth0      Link encap:Ethernet  HWaddr *06:88:14:00:00:8d  *
          inet addr:172.30.45.209  Bcast:172.30.45.255  Mask:255.255.255.0
          UP BROADCAST RUNNING MULTICAST  MTU:1500  Metric:1
          RX packets:190598 errors:0 dropped:0 overruns:0 frame:0
          TX packets:3959 errors:0 dropped:0 overruns:0 carrier:0
          collisions:0 txqueuelen:1000
          RX bytes:10158772 (9.6 MiB)  TX bytes:693574 (677.3 KiB)
          Interrupt:9

On the Host:
# xe vif-param-list uuid=47af9531-4303-8efc-edfd-c663e7068d07
uuid ( RO)                        : 47af9531-4303-8efc-edfd-c663e7068d07
                     vm-uuid ( RO): 3214aad8-b6dc-01ca-377c-47d737ba1056
               vm-name-label ( RO): r-4-VM
          allowed-operations (SRO): attach; unplug_force; unplug
          current-operations (SRO):
                      device ( RO): 0
                         MAC ( RO): *06:88:14:00:00:8d*
           MAC-autogenerated ( RO): false
                         MTU ( RO): 9000
          currently-attached ( RO): true
          qos_algorithm_type ( RW): ratelimit
        qos_algorithm_params (MRW): kbps: 25600
    qos_supported_algorithms (SRO):
                other-config (MRW): nicira-iface-id:
c38c93b3-b681-493b-9776-16ca857637d6; nicira-vm-id:
3214aad8-b6dc-01ca-377c-47d737ba1056
                network-uuid ( RO): *932c1b20-90d7-3d6e-a5c0-42dd555062aa*
          network-name-label ( RO): bond0
                 io_read_kbs ( RO): 0.097
                io_write_kbs ( RO): 0.000

# xe network-list uuid=*932c1b20-90d7-3d6e-a5c0-42dd555062aa*
uuid ( RO)                : 932c1b20-90d7-3d6e-a5c0-42dd555062aa
          name-label ( RW): bond0
    name-description ( RW):
              bridge ( RO): xapi1

# ifconfig xapi1
xapi1     Link encap:Ethernet  HWaddr d4:ae:52:b0:e3:53
          inet addr:172.30.45.31  Bcast:0.0.0.0  Mask:255.255.255.0
          inet6 addr: fe80::d6ae:52ff:feb0:e353/64 Scope:Link
          UP BROADCAST RUNNING MULTICAST  MTU:1500  Metric:1
          RX packets:1283441 errors:0 dropped:516 overruns:0 frame:0
          TX packets:788194 errors:0 dropped:0 overruns:0 carrier:0
          collisions:0 txqueuelen:0
          RX bytes:3439826848 (3.4 GB)  TX bytes:673693107 (673.6 MB)

Any ideas?

Host is Ubuntu 12.04 + default xapi-xcp

Other info:
The SSVM appears to be ok.  It is able to connect to secondary nfs storage
and I am able to connect to IPs outside.

I am not able to connect to my instance using the console from the
cloutstack UI.

The way I set up the network was the following:
#remove all the preconfigured xenbr bridges that xapi set up:
for i in `seq 0 3`; do
   ETH="eth$i"
   PIF=`xe pif-list device=$ETH --minimal`
   NET=`xe pif-param-get param-name=network-uuid uuid=${PIF}`
   xe network-destroy uuid=$NET
done
#create bond using eth0 and 1 and set it as the management network
ETH0=`xe pif-list device=eth0 --minimal`
ETH1=`xe pif-list device=eth1 --minimal`
NETID=`xe network-create name-label=bond0`
BONDID=`xe bond-create network-uuid=$NETID pif-uuids=$ETH0,$ETH1
mode=balance-slb`
BOND0=`xe pif-list device=bond0 --minimal`
xe pif-reconfigure-ip uuid=$BOND0 mode=static IP=172.30.45.31
netmask=255.255.255.0 gateway=172.30.45.1 DNS=172.30.20.176
xe host-management-reconfigure pif-uuid=$BOND0

I am wondering if the balance-slb is the problem.  Anyone have experience
with it?

The NFS primary storage network was setup outside of xapi with the os
configuration.