#2557·flannel

Carrier lost and dropped errors on flannel interface

Author: marek-obuchowiczCreated Sep 4, 2026Updated Sep 4, 2026

During higher traffic times, we've noticed that prometheus metric node_network_transmit_carrier_total increases on flannel0 interface. We're using xvlan mode on k8s clusters running in AWS, managed by kops. We also observe unexpected network issues regarding node-to-node, and node-to-external-service communication during periods where this counter increases (ie. Library error: a socket error occurred, or read error on connection to (...)).

# ip -s -s link show dev flannel.1
4: flannel.1: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 8951 qdisc noqueue state UNKNOWN mode DEFAULT group default
    link/ether a2:ae:54:71:22:58 brd ff:ff:ff:ff:ff:ff
    RX:   bytes  packets errors dropped  missed   mcast
     4412784279 11973699      0       0       0       0
    RX errors:    length    crc   frame    fifo overrun
                       0      0       0       0       0
    TX:   bytes  packets errors dropped carrier collsns
    28182728945 15251917      1      42       1       0
    TX errors:   aborted   fifo  window heartbt transns
                       0      0       0       0       0

There are no carrier/dropped packets on underlying physical interface:

# ip -s -s link show dev ens5
2: ens5: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 9001 qdisc mq state UP mode DEFAULT group default qlen 1000
    link/ether 0a:54:41:a1:fe:b5 brd ff:ff:ff:ff:ff:ff
    RX:   bytes  packets errors dropped  missed   mcast
    39702785904 38799104      0       0       0       0
    RX errors:    length    crc   frame    fifo overrun
                       0      0       0       0       0
    TX:   bytes  packets errors dropped carrier collsns
    37987895464 35810570      0       0       0       0
    TX errors:   aborted   fifo  window heartbt transns
                       0      0       0       0       3
    altname enp0s5