Skip site navigation (1)Skip section navigation (2)
Date:      Fri, 17 Jan 2014 10:50:12 +0100
From:      "Nagy, Attila" <bra@fsn.hu>
To:        FreeBSD Net <freebsd-net@freebsd.org>
Subject:   lagg member interfaces don't come back after cable pull
Message-ID:  <52D8FCD4.2040900@fsn.hu>

next in thread | raw e-mail | index | archive | help
Hi,

Running stable/9@r260621, multiple (currently two) igb's connected to 
multiple Cisco Nexus 2ks, forming an LACP port channel.

If I pull one of the cables from the switch, the port immediately goes 
down, and removed from the port channel. Everything works nicely.
When I reconnect the cable, the member interface (igb1) remains down.
On the host side, I see:
igb1: flags=8843<UP,BROADCAST,RUNNING,SIMPLEX,MULTICAST> metric 0 mtu 1500
options=401bb<RXCSUM,TXCSUM,VLAN_MTU,VLAN_HWTAGGING,JUMBO_MTU,VLAN_HWCSUM,TSO4,VLAN_HWTSO>
         ether ac:16:2d:9a:18:cd
         nd6 options=29<PERFORMNUD,IFDISABLED,AUTO_LINKLOCAL>
         media: Ethernet autoselect
         status: no carrier
igb2: flags=8843<UP,BROADCAST,RUNNING,SIMPLEX,MULTICAST> metric 0 mtu 1500
options=401bb<RXCSUM,TXCSUM,VLAN_MTU,VLAN_HWTAGGING,JUMBO_MTU,VLAN_HWCSUM,TSO4,VLAN_HWTSO>
         ether ac:16:2d:9a:18:cd
         nd6 options=29<PERFORMNUD,IFDISABLED,AUTO_LINKLOCAL>
         media: Ethernet autoselect (1000baseT <full-duplex>)
         status: active
admin: flags=8843<UP,BROADCAST,RUNNING,SIMPLEX,MULTICAST> metric 0 mtu 1500
options=401bb<RXCSUM,TXCSUM,VLAN_MTU,VLAN_HWTAGGING,JUMBO_MTU,VLAN_HWCSUM,TSO4,VLAN_HWTSO>
         ether ac:16:2d:9a:18:cd
         inet6 fe80::ae16:2dff:fe9a:18cd%admin prefixlen 64 tentative 
scopeid 0x9
         inet 172.27.79.74 netmask 0xffffff00 broadcast 172.27.79.255
         nd6 options=29<PERFORMNUD,IFDISABLED,AUTO_LINKLOCAL>
         media: Ethernet autoselect
         status: active
         laggproto lacp lagghash l2,l3,l4
         laggport: igb2 flags=1c<ACTIVE,COLLECTING,DISTRIBUTING>
         laggport: igb1 flags=0<>

Issuing an ifconfig down and up on igb1 fixes the problem, it gets back 
into the port channel.

On the switch side I can see the following:
2014 Jan 17 08:58:11 DP1106-A05-11-N5K %ETH_PORT_CHANNEL-5-PORT_UP: 
port-channel6: Ethernet109/1/47 is up
2014 Jan 17 09:20:42 DP1106-A05-11-N5K %ETH_PORT_CHANNEL-5-PORT_DOWN: 
port-channel6: Ethernet109/1/47 is down
2014 Jan 17 09:31:52 DP1106-A05-11-N5K %ETH_PORT_CHANNEL-5-PORT_UP: 
port-channel6: Ethernet109/1/47 is up

The cable was pulled at 09:20:42 and was reconnected in some seconds. 
ifconfig down and up was issued around 09:31:52, which fixed the access.

 From what I see it may be just the igb and not really connected to 
lagg, but I'm not sure and the machine is far away, so testing is not so 
easy.




Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?52D8FCD4.2040900>