Hello,
I'm using ifconfig to monitor the bandwidth usage of one of my VMs, but lately I've notified that while the VM has absolutely no traffic the TX number in ifconfig is counting up.

here it's the output of ifconfig on VM's interface in just 10 seconds interval:

vmtab302i0 Link encap:Ethernet HWaddr 36:c9:eb:02:02:49
inet6 addr: fe80::34c9:ebff:fe02:249/64 Scope:Link
UP BROADCAST RUNNING PROMISC MULTICAST MTU:1500 Metric:1
RX packets:430134 errors:0 dropped:0 overruns:0 frame:0
TX packets:999683632 errors:0 dropped:0 overruns:866628 carrier:0
collisions:0 txqueuelen:500
RX bytes:30844841 (29.4 MiB) TX bytes:136304055995 (126.9 GiB)

after 10 seconds:

vmtab302i0 Link encap:Ethernet HWaddr 36:c9:eb:02:02:49
inet6 addr: fe80::34c9:ebff:fe02:249/64 Scope:Link
UP BROADCAST RUNNING PROMISC MULTICAST MTU:1500 Metric:1
RX packets:430134 errors:0 dropped:0 overruns:0 frame:0
TX packets:999683870 errors:0 dropped:0 overruns:866628 carrier:0
collisions:0 txqueuelen:500
RX bytes:30844841 (29.4 MiB) TX bytes:136304077623 (126.9 GiB)


any suggestion about stopping this auto count up?

That's not the way to do things.


I'd say prove it using 'tcpdump' on guest and host. You might be missing ARP replies or whatever.

Almost all nics chatter in normal use. Use wireshark to see what the packets are.

I'm using iftop to watch the VMs traffic, and it has no traffic on it's IP.

Wireshark may be overkill if you're not going to use it's advanced features or don't run a GUI and iftop does not equal tcpdump.