Take the 2-minute tour ×
Unix & Linux Stack Exchange is a question and answer site for users of Linux, FreeBSD and other Un*x-like operating systems.. It's 100% free, no registration required.

I'm trying to configure a mode 6 bonding on a blade server.

  • Blade Enclosure: HP c7000
  • Blade Server: ProLiant BL460c G1

Redundancy part of the bond is working, I can pull out any one of the cables and the connection doesn't break.But for some strange reason the load balancing is not.The blade center has 2 network modules installed. Each network cable is connected to a different network module.I have another server (not blade) connected to the same switch and the bond is working great there, so it's not a switch issue.

Trying to figure it out for two days now, any help will be appreciated.

cat /proc/net/bonding/bond0
Ethernet Channel Bonding Driver: v3.7.1 (April 27, 2011)

Bonding Mode: adaptive load balancing
Primary Slave: None
Currently Active Slave: eth0
MII Status: up
MII Polling Interval (ms): 100
Up Delay (ms): 0
Down Delay (ms): 0

Slave Interface: eth0
MII Status: up
Speed: 1000 Mbps
Duplex: full
Link Failure Count: 0
Permanent HW addr: 00:21:5a:ce:39:b4
Slave queue ID: 0

Slave Interface: eth1
MII Status: up
Speed: 1000 Mbps
Duplex: full
Link Failure Count: 0
Permanent HW addr: 00:21:5a:ce:39:b2
Slave queue ID: 0
share|improve this question
    
Maybe this is because your are using mode 6(Adaptive Load Balance), where both NICs transmit data in a "mode 5 way(balance-tlb)" and it depends on the first NIC load to "delegate" traffic to the other bond slave, but only the active member of the moment will receive the traffic while the other changes it's source address to the bond mac address. Have your servers had enough load to start balancing things? –  nwildner Jun 10 '13 at 17:12
    
Yes they do. I start iperf server on the subject, and send data for 60 seconds from 5-6 other servers. This test works on other machines I have with bonding. –  KianTern Jun 11 '13 at 16:27
    
We had a problem a year ago with the balance-alb and 802.1q(we are not using OracleVM anymore) - docs.oracle.com/cd/E26996_01/E18546/html/… - the tests you are making with iperf are directly on the bond0 correct? Or it involves tagging and vlans? –  nwildner Jun 11 '13 at 18:03
    
I'm testing directly on bond0 –  KianTern Jun 12 '13 at 7:13
    
I'm pretty sure that this bug could be affecting directly ip related connections on the bridge. Is there any remote possibility to test another distro directly installed on this blade? –  nwildner Jun 12 '13 at 19:13
add comment

Your Answer

 
discard

By posting your answer, you agree to the privacy policy and terms of service.

Browse other questions tagged or ask your own question.