#18 Round Robin policy + arp monitoring (II)
(sorry I have just read that it's necessary to include a
valid email address)
joss_madrid@hotmail.com
Hello, I'm trying to agregate twoethernet interfaces on
two linux (redhat 7.4.20). I' using bonging (v2.4.20-
20030320"). I'd like not to use mii to monitor link
availability, so I'm trying with arp monitoring. Althought
when working in HA mode it works, when working on
mode 0 It doesn't work (it keeps on sending arp replies,
but the system doesn't care when the interface is not
receiving any frame, and losses one out of two packets)
----- -----
RH |----------- | RH |
|---------- | |
----- -----
mii-tool response
eth0: negotiated 100baseTx-FD, link ok
eth1: negotiated 100baseTx-FD, link ok
ethtool eth0 response
Settings for eth0:
Supported ports: [ TP MII ]
Supported link modes: 10baseT/Half 10baseT/Full
100baseT/Half 100baseT/Full
Supports auto-negotiation: Yes
Advertised link modes: 10baseT/Half 10baseT/Full
100baseT/Half 100baseT/Full
Advertised auto-negotiation: Yes
Speed: 100Mb/s
Duplex: Full
Port: MII
PHYAD: 32
Transceiver: internal
Auto-negotiation: on
Supports Wake-on: pumbg
Wake-on: d
Current message level: 0xffffffff (-1)
Link detected: yes
ethtool eth1
Supported ports: [ TP MII ]
Supported link modes: 10baseT/Half 10baseT/Full
100baseT/Half 100baseT/Full
Supports auto-negotiation: Yes
Advertised link modes: 10baseT/Half 10baseT/Full
100baseT/Half 100baseT/Full
Advertised auto-negotiation: Yes
Speed: 100Mb/s
Duplex: Full
Port: MII
PHYAD: 32
Transceiver: internal
Auto-negotiation: on
Supports Wake-on: pumbg
Wake-on: d
Current message level: 0xffffffff (-1)
Link detected: yes
Is it possible to make it work? maybe patching the
bonding driver?
Logged In: NO
I have a similar problem here:
two identical (cloned) gentoo boxes:
kernel: 2.6.9-rc1, gentoo-development
nics: one dualport Intel Corp. 82557/8/9 [Ethernet Pro 100],
driver used: e100
setup: PC A connects to two separate switches as does PC B
--------- sw1 --------
| |
A B
| |
--------- sw2 --------
bonding parameter: modprobe bonding arp_interval=2000
arp_ip_target=IP.OF.PC.B
bonding works as expected, but pulling one cable results in a slave
down status on one PC only!
Therefore every second packet gets lost.
This behaviour is the same with knoppix 3.3 on both PCs, running
a kernel version 2.4 as well as 2.6
All hardware components have been swapped to exclude hardware
problems.
What am I doing wrong?
Thanks in advance,
Florian Kessler
Logged In: NO
I have a similar problem here:
two identical (cloned) gentoo boxes:
kernel: 2.6.9-rc1, gentoo-development
nics: one dualport Intel Corp. 82557/8/9 [Ethernet Pro 100],
driver used: e100
setup: PC A connects to two separate switches as does PC B
--------- sw1 --------
| |
A B
| |
--------- sw2 --------
bonding parameter: modprobe bonding arp_interval=2000
arp_ip_target=IP.OF.PC.B
bonding works as expected, but pulling one cable results in a slave
down status on one PC only!
Therefore every second packet gets lost.
This behaviour is the same with knoppix 3.3 on both PCs, running
a kernel version 2.4 as well as 2.6
All hardware components have been swapped to exclude hardware
problems.
What am I doing wrong?
Thanks in advance,
Florian Kessler