Originally posted by: SystemAdmin

I've been fighting this one for a while, and haven't found much info at all regarding it.

The issue I'm having is that after a reboot the etherchannels that are defined only function on the primary connection (ie the first adapter in the list) and not any other adapters.
We have tried multiple switches, and every setting we can think of, all with the same result.

What does get the etherchannel working properly, is doing a detach, up, then mkdev on the etherchannel interface, ie;

ifconfig en6 detach

ifconfig en6 up

mkdev -l en6

Afterwhich the etherchannel begins to work as expected.

Doing a tcpdump, which puts the adapters in promisc mode, also gets things working.
Where as a

tcpdump -p

(which doesn't put it into promisc) shows that the etherchannel isn't receiving any traffic/replies (when the primary is down, until the above workaround is performed)

It seems as if when AIX initially brings up the etherchannel, it doesn't set something properly on the second adapter causing it to not accept any traffic. The only patches that google has come up with have been for versions earlier than what we're running, and I would expect they would be included in our version.

/tmp/rc.net.out does show this;

0821-510 ifconfig: error calling entry point 

for /usr/lib/drivers/if_en: The specified device does not exist. 0821-103 : The command /usr/sbin/ifconfig en6 inet  172.28.240.2 arp netmask 255.255.255.0 mtu 1500 up failed. 0821-007 cfgif: ifconfig command failed. The status of
"en6" Interface in the current running system is uncertain.

but according to, and this is about the only relevant doc I've been able to find so far, Link: developerworks:if_en: ns_alloc(en6) failed with errno 19 these errors are normal and shouldn't cause any issues. However, our etherchannels are only functioning on one adapater with any traffic being sent by the switch on the second adapter being ignored.
#AIX-Forum

Originally posted by: esv

execute the following command:

lsattr -El ent6

you should get something similar to:

prompt> lsattr -El ent2 adapter_names   ent0           EtherChannel Adapters                       True    <--- check 

this line alt_addr        0x000000000000 Alternate EtherChannel Address              True auto_recovery   yes            Enable automatic recovery after failover    True backup_adapter  ent1           Adapter used when whole channel fails       True    <--- check 

this line hash_mode       

default        Determines how outgoing adapter is chosen   True mode            standard       EtherChannel mode of operation              True netaddr         X.X.X.X        Address to ping                             True noloss_failover yes            Enable lossless failover after ping failure True num_retries     3              Times to retry ping before failing          True retry_time      1              Wait time (in seconds) between pings        True use_alt_addr    no             Enable Alternate EtherChannel Address       True use_jumbo_frame no             Enable Gigabit Ethernet Jumbo Frames        True prompt>

then execute the lsattr -El command for each interface which is part of the etherchannel...

lsattr -El en0 lsattr -El en1 etc...

if any of these devices has an IP address configured, then you'll have problems every single time, the best way I've found to get rid of these issues is using the chdev command on the standard ethernet network interfaces

chdev -l en0 -a netaddr=  -a netmask= -a state=down chdev -l en1 -a netaddr=  -a netmask= -a state=down

you might need to take the adapter off-line to make the change.
#AIX-Forum