1

I'm running Gentoo Hardened as Dom0. It also acts as a router/NAT. There's a bridge interface called br0 which contains a pci-e wlan card, ethernet nic & domU.

bridge name     bridge id               STP enabled     interfaces
br0             8000.002590d4dffe       no              enp5s0
                                                        vif1.0
                                                        wlp1s0

Internet connectiong goes through enp6s0 and masquerades to br0. Dnsmasq is serving addresses to br0. I'm experiencing a huge amount of dropped RX packets:

br0: flags=4419<UP,BROADCAST,RUNNING,PROMISC,MULTICAST>  mtu 1500
inet 192.168.0.1  netmask 255.255.255.0  broadcast 192.168.0.255
ether 00:25:90:d4:df:fe  txqueuelen 0  (Ethernet)
RX packets 61866962  bytes 67532965037 (62.8 GiB)
RX errors 0  dropped 26876  overruns 0  frame 0
TX packets 51716758  bytes 65149682646 (60.6 GiB)
TX errors 0  dropped 0 overruns 0  carrier 0  collisions 0

They also show up on domU:

eth0: flags=4163<UP,BROADCAST,RUNNING,MULTICAST>  mtu 1500
inet 192.168.0.2  netmask 255.255.255.0  broadcast 192.168.0.255
ether 00:16:3e:73:d7:5f  txqueuelen 1000  (Ethernet)
RX packets 4529002  bytes 4459840926 (4.1 GiB)
RX errors 0  dropped 35247  overruns 0  frame 0
TX packets 4487535  bytes 795673002 (758.8 MiB)
TX errors 0  dropped 0 overruns 0  carrier 0  collisions 0

I have no idea where to start debugging this. I've tried tcpdump and dropwatch. Here's output of ~10 seconds dropwatch run: http://pastebin.com/d0yUn9HD

ponkay
  • 11
  • 3

1 Answers1

0

Dropping packets is a sign of being overrun.

Look at ethtool -g eth0 and see if you can raise the receive-buffers.

Doing so will shortly (1-2s) reset the link.

Nils
  • 7,657
  • 3
  • 31
  • 71