Question about building a LAGG and odd behavior

R

rtbates63

Question about building a lagg(4) and odd behavior.

I have set up a system using FreeBSD 9.2. We are using a HP switch, ports B9, B10, B11 are LACP LAGG, port B12 is a standard port. On the server:
  • NIC em0 is connected to port B12, IP address 172.18.1.134
  • NICs em1,em4,em5 are connected to HP ports B9, B10, B11, IP address 172.18.1.138

In /etc/rc.conf we have:
Code: 
ifconfig_em0="inet 172.18.1.134 netmask 255.255.255.0"
ifconfig_em1="up polling"
ifconfig_em4="up polling"
ifconfig_em5="up polling"
cloned_interfaces="lagg0"
ifconfig_lagg0="up laggproto lacp laggport em1 laggport em4 laggport em5 172.18.1.138/24"
defaultrouter="172.18.1.2"

When it seems to work. But when I disconnect the single CAT5 172.18.1.134 the lagg0 goes down and I can not ping the server on 172.18.1.138. Also, I noticed on the HP switch most of the traffic is going over server port em0 only a small portion of traffic is going over the lagg interface. (See HP Port Usage image attached).

Can you help me with this configuration?

Thank you.
 

Attachments

  • HP port Usage.jpg
    HP port Usage.jpg
    12.2 KB · Views: 390
You have two interfaces, em0 and lagg0, in the same subnet. That's never a good idea.
 
Yes, I know. But when I remove the IP off of em0, I lose the connection to the LAGG. That's the odd behavior. Couldn't get LAGG to come up without adding the one extra direct connect port.

Also, I originally had the em0 defined so I could finish config from my desk (not the server room). Once I configured LAGG I attempted to remove the em0 port and reboot. LAGG never came up. Had to physically go to the server room and configure the em0.
 
You must log in or register to reply here.
Back
Top