Hi All,
We've got an on premises server installed on our core switch (Stacked).
When I tested it on a different rack switch (single), the slave was joining to the bond0 (Excuse my knowledge with the server configuration as I don't really know much about this and didn't configure it).
This test area port channel is configured to LACP mode as below.
interface Port-channel19
description * SERVER SUKDEB TESTING *
switchport access vlan 1112
switchport mode access
spanning-tree portfast
interface GigabitEthernet1/0/21
description * SERVER SUKDEB TESTING *
switchport access vlan 1112
switchport mode access
channel-group 19 mode active
spanning-tree portfast
!
interface GigabitEthernet1/0/22
description * SERVER SUKDEB TESTING *
switchport access vlan 1112
switchport mode access
channel-group 19 mode active
spanning-tree portfast
From the etherchannel summary, the port channel 19 is up and running.
19 Po19(SU) LACP Gi1/0/21(P) Gi1/0/22(P)
The server configuration guy said LAG is working on his end.
On the core switch, the port channel is set to "ON" mode as it's just server at the end and not connected to anything else.
Also the port channel was not going live with "Active".
Here's the configuration on the core switch.
interface Port-channel20
description * SERVER MAIN COMMS *
switchport trunk allowed vlan 1008,1112
switchport mode trunk
!
interface GigabitEthernet2/0/10
description *SERVER MAIN COMMS*
switchport trunk allowed vlan 1008,1112
switchport mode trunk
channel-group 20 mode on
!
interface GigabitEthernet3/0/10
description *SERVER MAIN COMMS*
switchport trunk allowed vlan 1008,1112
switchport mode trunk
channel-group 20 mode on
Etherchannel summary shows it's up and running.
20 Po20(SU) - Gi2/0/10(P) Gi3/0/10(P)
This is my side of the story.
And here's the server side of story.
The configuration is just the same as our local server, should work just fine as local one with core switch if switch side configuration is identical. I went there and checked that both links are blinking green and instead of 4, Edwards suggested to use two links on interface eno3 and eno4.
To take both to NEP right now we need to:-
1. we need to configure 8 total ports from switch to support LACP protocol.
2. Two different static IP for those two RAC servers, one has got 10.31.8.200, OS inside it has 10.31.8.210 please see the following, it says both links up means connected, green blinking but link failure coute is 1, has to be 0.
And no slaves are joining the master bond0 as the ports are not supporting LACP.
bond0 status in kernel filesystem:-
Do you guys have any opinion why the server is not working with the port channel?
Does it need to be LACP? If so, is there a way to change this to
