Bug 2172578

Reproduce this bug on libvirt-8.0.0-17.module+el8.8.0+18133+90800a0a.x86_64
# ip tuntap add mode tap name mytap0
# ip link set mytap0 up
# ip link set mytap0 master br0
# virsh dumpxml rhel | grep /interface -B6
   <interface type='ethernet'>
     <target dev='mytap0' managed='no'/>
     <model type='virtio'/>
   </interface>

Login guest, and run:
# curl -O https://download.fedoraproject.org/pub/fedora/linux/releases/37/Cloud/x86_64/images/Fedora-Cloud-Base-37-1.7.x86_64.qcow2 -L

Check the statistics on guest and host:
On guest:
# ip -s l show enp1s0
2: enp1s0: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 1500 qdisc fq_codel state UP mode DEFAULT group default qlen 1000
    link/ether 52:54:00:88:8b:23 brd ff:ff:ff:ff:ff:ff
    RX:  bytes packets errors dropped  missed   mcast           
     517891022  357101      0     225       0       0 
    TX:  bytes packets errors dropped carrier collsns           
      20557310  253394      0       0       0       0

On host:
# virsh domifstat rhel mytap0
mytap0 rx_bytes 20557310
mytap0 rx_packets 253394
mytap0 rx_errs 0
mytap0 rx_drop 0
mytap0 tx_bytes 517891112
mytap0 tx_packets 357102
mytap0 tx_errs 0
mytap0 tx_drop 0
The domifstat command shows the statistics from the host view, which is unexpected. 

While with bridge type interface as below, the domifstat command shows the statistics from the guest point of view, which is as expected.
# virsh dumpxml rhel | grep /interface -B7
    <interface type='bridge'>
      <mac address='52:54:00:41:7c:2c'/>
      <source bridge='br0'/>
      <target dev='vnet1'/>
      <model type='virtio'/>
      <alias name='net0'/>
      <address type='pci' domain='0x0000' bus='0x01' slot='0x00' function='0x0'/>
    </interface>

Download the file on guest, then check the statistics on host and guest:
On host:
# virsh domifstat rhel vnet1
vnet1 rx_bytes 509510064
vnet1 rx_packets 227805
vnet1 rx_errs 0
vnet1 rx_drop 0
vnet1 tx_bytes 14086386
vnet1 tx_packets 178018
vnet1 tx_errs 0
vnet1 tx_drop 0

On guest:
# ip -s l show enp1s0
2: enp1s0: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 1500 qdisc fq_codel state UP mode DEFAULT group default qlen 1000
    link/ether 52:54:00:41:7c:2c brd ff:ff:ff:ff:ff:ff
    RX:  bytes packets errors dropped  missed   mcast           
     509509914  227803      0    1155       0       0 
    TX:  bytes packets errors dropped carrier collsns           
      14086386  178018

From the test result above, the "virsh domifstat" can not catch the RX_dropped value, is that acceptable? 

On host, I have checked the rx_dropped file, it's also 0. 
# cat /sys/devices/virtual/net/vnet1/statistics/rx_dropped 
0

(In reply to yalzhang from comment #4)
> From the test result above, the "virsh domifstat" can not catch the
> RX_dropped value, is that acceptable? 
> 
> On host, I have checked the rx_dropped file, it's also 0. 
> # cat /sys/devices/virtual/net/vnet1/statistics/rx_dropped 
> 0

Yes it is. Libivirt does nothing more than read this value. If kernel reports it as zero there's not much we can do.

Test on libvirt-8.0.0-18.module+el8.8.0+18287+112bb4e6.x86_64 with the scenarios in bz2172574#c4  bz2172574#c5, the result is as expected.

Since the problem described in this bug report should be
resolved in a recent advisory, it has been closed with a
resolution of ERRATA.

For information on the advisory (Moderate: virt:rhel and virt-devel:rhel security, bug fix, and enhancement update), and where to find the updated
files, follow the link below.

If the solution does not work for you, open a new bug report.

https://access.redhat.com/errata/RHSA-2023:2757