Re: [ovs-discuss] Issue when using local_ip with VXLAN tunnels in OVS

Gregory Rose Tue, 20 Nov 2018 09:26:26 -0800


On 11/19/2018 6:30 PM, Siva Teja ARETI wrote:

On Mon, Nov 19, 2018 at 7:17 PM Gregory Rose <gvrose8...@gmail.com<mailto:gvrose8...@gmail.com>> wrote:
    Hi Siva,

    One more request  - I need to see the underlying network
    configuration
    of the hypervisor running the two VMs.
    Are both VMs on the same machine?  If so then just the network
    configuration of the base machine
    running the VMs, otherwise the network configuration of each base
    machine running their perspective
    VM.

    This is turning into quite the investigation and I apologize that
    it is
    taking so long.  Please bear with me
    if you can and we'll see if we can't get this problem solved. 
    I've seen
    some puzzling bugs before and
    this one is turning out to be one of the best.  Or worst....
    depends on
    your outlook.  :)

    Thanks for all your help so far!

    - Greg


Hi Greg,
Both the VMs run on same hypervisor in my setup. Created VMs andvirtual networks using virsh commands. Virsh XMLs for networks looklike below
[user@hyp1 ] virsh net-dumpxml route1
<network connections='2'>
  <name>route1</name>
<uuid>2c935aaf-ebde-5b76-a903-4fccb115ff75</uuid>
  <forward mode='route'/>
  <bridge name='testbr1' stp='on' delay='0'/>
  <mac address='42:54:00:84:4e:04'/>
  <ip address='20.20.0.1' netmask='255.255.255.0'>
    <dhcp>
      <range start='20.20.0.2' end='20.20.0.254'/>
    </dhcp>
  </ip>
</network>

[user@hyp1 ]  network virsh net-dumpxml route2
<network connections='2'>
  <name>route2</name>
<uuid>2c935baf-ebde-5b76-a903-4fccb115ff75</uuid>
  <forward mode='route'/>
  <bridge name='testbr2' stp='on' delay='0'/>
  <mac address='42:54:10:84:4e:04'/>
  <ip address='30.30.0.1' netmask='255.255.255.0'>
    <dhcp>
      <range start='30.30.0.2' end='30.30.0.254'/>
    </dhcp>
  </ip>
</network>

Each VM is connected to both the networks.

Some network configuration of the hypervisor.

[user@hyp-1] ip a
1: lo: <LOOPBACK,UP,LOWER_UP> mtu 65536 qdisc noqueue state UNKNOWNgroup default qlen 1
    link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00
    inet 127.0.0.1/8 <http://127.0.0.1/8> scope host lo
       valid_lft forever preferred_lft forever
    inet6 ::1/128 scope host
       valid_lft forever preferred_lft forever
2: enp5s0: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 1400 qdisc pfifo_faststate UP group default qlen 1000
    link/ether <mac> brd ff:ff:ff:ff:ff:ff
    inet A.B.C.D/24 brd X.Y.Z.W scope global dynamic enp5s0
       valid_lft 318349sec preferred_lft 318349sec
3: virbr0: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 1500 qdisc noqueuestate UP group default qlen 1000
    link/ether fe:54:00:0a:d3:70 brd ff:ff:ff:ff:ff:ff
inet 192.168.122.1/24 <http://192.168.122.1/24> brd192.168.122.255 scope global virbr0
       valid_lft forever preferred_lft forever
4: virbr0-nic: <BROADCAST,MULTICAST> mtu 1500 qdisc pfifo_fast stateDOWN group default qlen 1000
    link/ether 52:54:00:94:4e:04 brd ff:ff:ff:ff:ff:ff
11: docker0: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 1500 qdisc noqueuestate UP group default
    link/ether 02:42:89:28:db:a5 brd ff:ff:ff:ff:ff:ff
    inet 172.17.0.1/16 <http://172.17.0.1/16> scope global docker0
       valid_lft forever preferred_lft forever
    inet6 fe80::42:89ff:fe28:dba5/64 scope link
       valid_lft forever preferred_lft forever
96: vboxnet0: <NO-CARRIER,BROADCAST,MULTICAST,UP> mtu 1500 qdiscpfifo_fast state DOWN group default qlen 1000
    link/ether 0a:00:27:00:00:00 brd ff:ff:ff:ff:ff:ff
inet 192.168.99.1/24 <http://192.168.99.1/24> brd 192.168.99.255scope global vboxnet0
       valid_lft forever preferred_lft forever
    inet6 fe80::800:27ff:fe00:0/64 scope link
       valid_lft forever preferred_lft forever
193: testbr0: <NO-CARRIER,BROADCAST,MULTICAST,UP> mtu 1500 qdiscnoqueue state DOWN group default qlen 1000
    link/ether 00:00:00:00:00:00 brd ff:ff:ff:ff:ff:ff
inet 10.10.0.1/24 <http://10.10.0.1/24> brd 10.10.0.255 scopeglobal testbr0
       valid_lft forever preferred_lft forever
194: testbr0-nic: <BROADCAST,MULTICAST> mtu 1500 qdisc pfifo_faststate DOWN group default qlen 1000
    link/ether 42:54:00:94:4e:04 brd ff:ff:ff:ff:ff:ff
227: testbr1: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 1500 qdisc noqueuestate UP group default qlen 1000
    link/ether fe:54:00:05:93:7c brd ff:ff:ff:ff:ff:ff
inet 20.20.0.1/24 <http://20.20.0.1/24> brd 20.20.0.255 scopeglobal testbr1
       valid_lft forever preferred_lft forever
228: testbr1-nic: <BROADCAST,MULTICAST> mtu 1500 qdisc pfifo_faststate DOWN group default qlen 1000
    link/ether 42:54:00:84:4e:04 brd ff:ff:ff:ff:ff:ff
229: testbr2: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 1500 qdisc noqueuestate UP group default qlen 1000
    link/ether fe:54:00:79:ef:92 brd ff:ff:ff:ff:ff:ff
inet 30.30.0.1/24 <http://30.30.0.1/24> brd 30.30.0.255 scopeglobal testbr2
       valid_lft forever preferred_lft forever
230: testbr2-nic: <BROADCAST,MULTICAST> mtu 1500 qdisc pfifo_faststate DOWN group default qlen 1000
    link/ether 42:54:10:84:4e:04 brd ff:ff:ff:ff:ff:ff
231: vnet0: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 1500 qdiscpfifo_fast master virbr0 state UNKNOWN group default qlen 1000
    link/ether fe:54:00:0a:d3:70 brd ff:ff:ff:ff:ff:ff
    inet6 fe80::fc54:ff:fe0a:d370/64 scope link
       valid_lft forever preferred_lft forever
232: vnet1: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 1500 qdiscpfifo_fast master testbr2 state UNKNOWN group default qlen 1000
    link/ether fe:54:00:b8:05:be brd ff:ff:ff:ff:ff:ff
    inet6 fe80::fc54:ff:feb8:5be/64 scope link
       valid_lft forever preferred_lft forever
233: vnet2: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 1500 qdiscpfifo_fast master testbr1 state UNKNOWN group default qlen 1000
    link/ether fe:54:00:f0:64:37 brd ff:ff:ff:ff:ff:ff
    inet6 fe80::fc54:ff:fef0:6437/64 scope link
       valid_lft forever preferred_lft forever
234: vnet3: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 1500 qdiscpfifo_fast master virbr0 state UNKNOWN group default qlen 1000
    link/ether fe:54:00:56:cb:89 brd ff:ff:ff:ff:ff:ff
    inet6 fe80::fc54:ff:fe56:cb89/64 scope link
       valid_lft forever preferred_lft forever
235: vnet4: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 1500 qdiscpfifo_fast master testbr2 state UNKNOWN group default qlen 1000
    link/ether fe:54:00:79:ef:92 brd ff:ff:ff:ff:ff:ff
    inet6 fe80::fc54:ff:fe79:ef92/64 scope link
       valid_lft forever preferred_lft forever
236: vnet5: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 1500 qdiscpfifo_fast master testbr1 state UNKNOWN group default qlen 1000
    link/ether fe:54:00:05:93:7c brd ff:ff:ff:ff:ff:ff
    inet6 fe80::fc54:ff:fe05:937c/64 scope link
       valid_lft forever preferred_lft forever
[user@hyp-1] ip route
default via A.B.C.D dev enp5s0  proto static  metric 100
10.10.0.0/24 <http://10.10.0.0/24> dev testbr0 proto kernel scopelink src 10.10.0.1 linkdown20.20.0.0/24 <http://20.20.0.0/24> dev testbr1 proto kernel scopelink src 20.20.0.130.30.0.0/24 <http://30.30.0.0/24> dev testbr2 proto kernel scopelink src 30.30.0.1
A.B.C.D via P.Q.R.S dev enp5s0  proto dhcp  metric 100
X.Y.Z.W dev enp5s0  proto kernel  scope link  src A.B.C.D  metric 100
172.17.0.0/16 <http://172.17.0.0/16> dev docker0 proto kernel scopelink src 172.17.0.1192.168.99.0/24 <http://192.168.99.0/24> dev vboxnet0 proto kernel scope link src 192.168.99.1 linkdown192.168.122.0/24 <http://192.168.122.0/24> dev virbr0 proto kernel scope link src 192.168.122.1
I am not completely sure what you meant by network configuration.Kindly let me know if you are looking for something more specific.
There is one strange behavior that I observed on the VMs. I am able toping across networks if I use IP address but it does not work if I useinterface name directly.
[root@vm2 ~]# ip addr  show eth1
3: eth1: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 1500 qdisc pfifo_faststate UP qlen 1000
    link/ether 52:54:00:79:ef:92 brd ff:ff:ff:ff:ff:ff
inet 30.30.0.193/24 <http://30.30.0.193/24> brd 30.30.0.255 scopeglobal dynamic eth1
       valid_lft 2728sec preferred_lft 2728sec
    inet6 fe80::5054:ff:fe79:ef92/64 scope link
       valid_lft forever preferred_lft forever
[root@vm2 ~]# ping 20.20.0.183 -I eth1
PING 20.20.0.183 (20.20.0.183) from 30.30.0.193 eth1: 56(84) bytes ofdata.
^C
--- 20.20.0.183 ping statistics ---
7 packets transmitted, 0 received, 100% packet loss, time 6000ms

[root@vm2 ~]# ping 20.20.0.183 -I 30.30.0.193
PING 20.20.0.183 (20.20.0.183) from 30.30.0.193 : 56(84) bytes of data.
64 bytes from 20.20.0.183 <http://20.20.0.183>: icmp_seq=1 ttl=64time=0.766 ms64 bytes from 20.20.0.183 <http://20.20.0.183>: icmp_seq=2 ttl=64time=0.561 ms64 bytes from 20.20.0.183 <http://20.20.0.183>: icmp_seq=3 ttl=64time=0.605 ms64 bytes from 20.20.0.183 <http://20.20.0.183>: icmp_seq=4 ttl=64time=0.537 ms64 bytes from 20.20.0.183 <http://20.20.0.183>: icmp_seq=5 ttl=64time=0.607 ms64 bytes from 20.20.0.183 <http://20.20.0.183>: icmp_seq=6 ttl=64time=0.618 ms64 bytes from 20.20.0.183 <http://20.20.0.183>: icmp_seq=7 ttl=64time=0.624 ms
^C
--- 20.20.0.183 ping statistics ---
7 packets transmitted, 7 received, 0% packet loss, time 6000ms
rtt min/avg/max/mdev = 0.537/0.616/0.766/0.075 ms
[root@vm2 ~]#
I don't the reason behind this and I will need to understand this whenI get some time. Just letting you know if it might make any differencein your setup.
Siva Teja.


Siva,

Thanks for that information - it will help. I'll have to spend sometime analyzing it, this is a complex setup.

As for the ping - that is to be expected. The IP address belongs to thesystem so any IP address on the systemcan be pinged. That's why I mentioned adding the '-I' option to specifythe interface, that way you force theping through a specific interface and that will help understand therouting setup on a machine.


Thanks,

- Greg

_______________________________________________
discuss mailing list
disc...@openvswitch.org
https://mail.openvswitch.org/mailman/listinfo/ovs-discuss

Re: [ovs-discuss] Issue when using local_ip with VXLAN tunnels in OVS

Reply via email to