Sponsored Content
Full Discussion: RHEL 5.5 NIC Bonding
Operating Systems Linux Red Hat RHEL 5.5 NIC Bonding Post 302768365 by tjcooper on Friday 8th of February 2013 12:00:18 PM
Old 02-08-2013
RHEL 5.5 NIC Bonding

I am new to linux. I have a falconstor CDP installation which runs on RHEL 5.5 (Tikanga). There is a vendor option to bond nics in either 802.3ad or round robin, I choose 802.3ad. The server is connected to a Cisco 3750 switch.

The problem is dropped packets and eth03 wont come up in the PO (suspended). I am getting reports of dropped packets. When I look at the network files, things dont look right. I am not posting this without doing homework, I have spent the last couple days looking into RHEL NIC Bonding. At this point, I need help, before I make the changes I think are necessary. I dont think Eth0 should have an IP and I am not entirely sure what kind of load-balancing RHEL supports.

This is what is on the server

Code:
DEVICE=bond0
IPADDR=192.168.0.27
NETMASK=255.255.255.0
USERCTL=no
BOOTPROTO=none
ONBOOT=yes
MTU=1500
BONDING_OPTS="mode=4 miimon=100 updelay=60000"
 
DEVICE=eth0
ONBOOT=yes
BOOTPROTO=none
#IPADDR=192.168.0.27
#NETMASK=255.255.255.0
HWADDR=00:23:7d:e1:f8:66
DHCP_HOSTNAME=
MTU=1500
MASTER=bond0
SLAVE=yes
 
DEVICE=eth1
ONBOOT=yes
BOOTPROTO=none
#IPADDR=10.0.0.3
#NETMASK=255.255.255.0
TYPE=Ethernet
HWADDR=00:23:7d:e1:f8:68
DHCP_HOSTNAME=
MTU=1500
MASTER=bond0
SLAVE=yes
 
DEVICE=eth2
ONBOOT=yes
BOOTPROTO=none
#IPADDR=10.0.0.4
#NETMASK=255.255.255.0
TYPE=Ethernet
HWADDR=00:23:7d:e1:f8:6a
DHCP_HOSTNAME=
MTU=1500
MASTER=bond0
SLAVE=yes
 
DEVICE=eth3
ONBOOT=yes
BOOTPROTO=none
#IPADDR=10.0.0.5
#NETMASK=255.255.255.0
TYPE=Ethernet
HWADDR=00:23:7d:e1:f8:6c
DHCP_HOSTNAME=
MTU=1500
MASTER=bond0
SLAVE=yes
 
Modprobe.config
alias bond0 bonding
alias eth0 bnx2
alias eth1 bnx2
alias eth2 bnx2
alias eth3 bnx2
alias scsi_hostadapter cciss
alias scsi_hostadapter1 libata
alias scsi_hostadapter2 ata_piix
options scsi_mod max_luns=1
options sd_mod max_scsi_disks_fs=1
options bonding max_bonds=1
options bond0 miimon=100 mode=4 updelay=60000

CISCO SWITCH

*each port looks like this

Code:
interface GigabitEthernet1/0/5
 switchport access vlan 500
 switchport mode access
 switchport nonegotiate
 channel-group 1 mode active
 spanning-tree portfast
end 
*load-balance
 
SCHQYADR3750#show etherchannel load-balance
EtherChannel Load-Balancing Operational State (src-mac):
Non-IP: Source MAC address
  IPv4: Source MAC address
  IPv6: Source IP address


Last edited by DukeNuke2; 02-08-2013 at 02:02 PM..
 

9 More Discussions You Might Find Interesting

1. Red Hat

External NIC cards not pinging with each other in RHEL 5.1

We have a setup of two nodes which has two NIC cards in each. One is built in and another is D-Link (external NIC card). We are unable to ping to the external NIC cards in both. Eg: Node A has two IPs 192.168.1.10 (eth0) and 192.168.2.10 (eth1) and Node B has two IPs 192.168.1.20 (eth0) and... (3 Replies)
Discussion started by: durgaprasadr13
3 Replies

2. Red Hat

Does RHEL 5 provide a command to collect RHEL system log in single compress file?

Hi, I heard a command that can collect all RHEL 5 log in a single compress file before I forget. Does any body know...What the command is ? Thanks. (4 Replies)
Discussion started by: nnnnnnine
4 Replies

3. Solaris

x86 Solaris 10 nic driver added but not attached. NIC is not detected.

I couldn't install my nic in solaris 10. I compiled and added the driver but failed to attach the driver and ifconfig output shows only loopback dev. Please see the following output and tell me whether my nic has been detected and why the driver failed to attach? My nic is detected in linux... (0 Replies)
Discussion started by: vectrum
0 Replies

4. IP Networking

squid proxy: one NIC for inbound & one NIC for outbound?

I am new in squid proxy. My question is how to (and if it's necessary) to set one NIC for inbound traffic (http requests) and one NIC for outbound traffic (http answers)? Thank you in advance! (4 Replies)
Discussion started by: aixlover
4 Replies

5. Red Hat

I want to tune NIC's rps, rfs and xps value. which NIC device should I modify.

Dear All I want tune my NIC's rps, rfs and xps value. In my system I have two NIC (eth0, eth1) and I have a bond0 ( eth0, eth1). Here is the question? Which device should I modify ? eth0 and eth1? or just modify bond0 or modify all device (eth0, eth1, bond0) Any advice is welcome.... (0 Replies)
Discussion started by: nnnnnnine
0 Replies

6. Red Hat

Error throwing while installing vsftpd package in rhel 6. using rhel 6 dvd.

Hi all, Im studying rhcsa as of now, so yum installation and dependencies are messing me to not workit out. i have dual os, win 7 & rhel 6. i have tried this installation of vsftpd package with rhel 6 dvd in VM rhel 6 in win 7 as well as host rhel 6.still the same issue. below error... (6 Replies)
Discussion started by: redhatlbug
6 Replies

7. IP Networking

Is bond0 ready and a bonding interface? - 4G Bonding Router Problems (RPI)

HI all, First post on the forum, and my first proper project on the Paspberry Pi, so sorry if this is in the wrong place. I am trying to turn my Pi in to a 3G/4G Bonding router. I have been researching and trying this for a week or so now. The basic plan is to have up to 6 ZTE MF823 USB... (0 Replies)
Discussion started by: elliottlc
0 Replies

8. Red Hat

RHEL 6, Spacewalk 2.3 unable to download RHEL 5 repo data

Hello all, I am having a bit of an issue on my Spacewalk installation. Some amplifying information is that it is Spacewalk 2.3 installed on a RHEL 6 machine and I am attempting to install/update a RHEL 5 channel/repository. I am fairly new to Spacewalk so I am still learning but this is what I... (3 Replies)
Discussion started by: jstone4646
3 Replies

9. Linux

Add two different subnet public IPs to single NIC or two different NIC on same box

Hello Admins, My ask is how can I add two different subnet IPs to same box with two different gateways? The issue is I can connect to the box when I am on ethernet LAN, but I am not able to connect to the same IP when I am on wifi. The server is RHEL 7 VM on vmware. How can I get connected... (4 Replies)
Discussion started by: snchaudhari2
4 Replies
AGR(4)							   BSD Kernel Interfaces Manual 						    AGR(4)

NAME
agr -- link aggregation pseudo network interface driver SYNOPSIS
pseudo-device agr DESCRIPTION
The agr driver provides link aggregation functionality (a.k.a. L2 trunking or bonding). It supports the IEEE 802.3ad Link Aggregation Control Protocol (LACP) and the Marker Protocol. The agr driver supports the following link specific flags for ifconfig(8): link0 Use the round-robin distribution algorithm. Don't use it unless you're really sure, because it violates the frame ordering rule. -link0 Use the default distribution algorithm, which is based on the hash of DA/SA, TCI, and, if available, some upper layer protocol infor- mation like ip(4) DA/SA. link1 Disable LACP. Prevents any LACP or Marker messaging which leaves the ports in the default static configuration. Set this prior to adding ports. EXAMPLES
Create an agr interface, agr0, and attach re0 and re1 to it. In other words, aggregate re0 and re1 so that they can be used as a single interface, agr0. The physical interfaces which are attached to the agr interface must not have any IP addresses, neither IPv4 nor IPv6. ifconfig re0 inet xxx.xxx.xxx.xxx delete ifconfig re0 inet6 fe80::xxxx:xxxx:xxxx:xxxx delete ifconfig re1 inet xxx.xxx.xxx.xxx delete ifconfig re1 inet6 fe80::xxxx:xxxx:xxxx:xxxx delete ifconfig agr0 create ifconfig agr0 agrport re0 ifconfig agr0 agrport re1 Destroy an interface created in the above example. ifconfig agr0 -agrport re0 ifconfig agr0 -agrport re1 ifconfig agr0 destroy SEE ALSO
ifconfig(8) STANDARDS
IEEE 802.3ad Aggregation of Multiple Link Segments HISTORY
The agr driver first appeared in NetBSD 4.0. AUTHORS
The agr driver was written by YAMAMOTO Takashi. BUGS
There is no way to configure LACP administrative variables, including system and port priorities. The current implementation of the agr driver always performs active-mode LACP and uses 0x8000 as system and port priorities. The agr driver uses the MAC address of the first-added physical interface as the MAC address of the agr interface itself. Thus, removing the physical interface and using it for another purpose can result in non-unique MAC addresses. The current implementation of the agr driver doesn't prevent unsafe operations like some ioctls against underlying physical interfaces. Such operations can result in unexpected behaviors, and are strongly discouraged. There is no way to configure agr interfaces without attaching physical interfaces. Physical interfaces being added to the agr interface shouldn't have any addresses except for link level address. Otherwise, the attempt will fail with EBUSY. Note that it includes an automatically assigned IPv6 link-local address. BSD
February 23, 2010 BSD
All times are GMT -4. The time now is 08:17 AM.
Unix & Linux Forums Content Copyright 1993-2022. All Rights Reserved.
Privacy Policy