Problem with etherchannel on a Cisco 2960S

[netgus2]

Hello everybody,

I need to make a etherchannel because a don't have a lot of ethernet card.

This is my configuration and a some "show command" to help you to identify, I hope so, my problem.

C2960S#sh etherchannel sum
Flags:  D - down         P - bundled in port-channel
     I - stand-alone s - suspended
     H - Hot-standby (LACP only)
     R - Layer3       S - Layer2
     U - in use       f - failed to allocate aggregator

     M - not in use, minimum links not met
     u - unsuitable for bundling
     w - waiting to be aggregated
     d - default port


Number of channel-groups in use: 2
Number of aggregators:             2

Group  Port-channel  Protocol    Ports
------+-------------+-----------+-----------------------------------------------
1       Po1(SU)           LACP       Gi1/0/27(P) Gi1/0/28(P) (This is a etherchannel between my 2960S and my 3750G. This 3750G is VTP server and a 2960S is VTP client and all VLAN are declared)
2       Po2(SU)            -         Gi1/0/22(P) Gi1/0/23(P) Gi1/0/24(P) (This, This is the etherchannel between my ESXI v5 server)

If I do a sh int trunk, this trunk It's ok:

C2960S#sh int trunk

Port         Mode                Encapsulation  Status         Native vlan
Po1           on                  802.1q           trunking       1
Po2           on                  802.1q           trunking       1

Port         Vlans allowed on trunk
Po1           1-4094
Po2           1,4,6,100

Port         Vlans allowed and active in management domain
Po1           1,3-7,10-12,50,99-100,801
Po2           1,4,6,100

Port         Vlans in spanning tree forwarding state and not pruned
Po1           1,3-7,10-12,50,99-100,801
Po2           1,4,6,100

This is my cisco 2960 configuration:
interface Port-channel2
description ESX_VLAN_TRUNK_2
switchport trunk allowed vlan 1,4,6,100
switchport mode trunk
switchport nonegotiate

interface GigabitEthernet1/0/22
switchport trunk allowed vlan 1,4,6,100
switchport mode trunk
switchport nonegotiate
speed 1000
duplex full
channel-group 2 mode on
!
interface GigabitEthernet1/0/23
switchport trunk allowed vlan 1,4,6,100
switchport mode trunk
switchport nonegotiate
speed 1000
duplex full
channel-group 2 mode on
!
interface GigabitEthernet1/0/24
switchport trunk allowed vlan 1,4,6,100
switchport mode trunk
switchport nonegotiate
speed 1000
duplex full
channel-group 2 mode on

Other show:

Port-channel: Po2
------------

Age of the Port-channel   = 0d:02h:43m:04s
Logical slot/port   = 5/2            Number of ports = 1
GC                      = 0x00000000       HotStandBy port = null
Port state            = Port-channel Ag-Inuse
Protocol              =    -
Port security        = Disabled

Ports in the Port-channel:

Index   Load   Port      EC state         No of bits
------+------+------+------------------+-----------
   0      00      Gi1/0/21 On                     0

Time since last port bundled:    0d:00h:02m:48s    Gi1/0/21
Time since last port Un-bundled: 0d:00h:02m:51s    Gi1/0/21

port-channel load-balance src-dst-ip is configured on my Cisco 2960S-G

After the configuration has been placed, it can not ping or access to the console with the administration IP address.
The only way is to shutdown all interfaces except one and then you can access the console. If after I do a no sh in another interface, the communication down immediatly.

VMWare,
- Load balancing is set to "Route based on IP hash", "State links", "yes", "yes"
- All adapters are reported as active
- Full duplex & speed to 1000 for each adapter
- The VLANID 6 is said to "Management Network"
- A port group server VLANID100
- Etc. ..

I admit not knowing what to do.

If everyone can help me, it'll be fine.

Thank

[abirhasan]

Here is a reference that may help ..

http://www.cisco.com/en/US/products/hw/switches/ps628/products_configuration_example09186a00800ef797.shtml

abirhasan 

[spravtek]

Hi ...

It's a little difficult to see what's going on with only some switch information, it would be beneficial to have an idea about the physical setup and maybe some screenshots of the network configuration in either vCenter or the ESXi host ...

Did you have a read through the KB with etherchannel examples?

http://kb.vmware.com/selfservice/microsites/search.do?language=en_US&cmd=displayKC&externalId=1004048

[netgus2]

Thank you for our answer,

I put some pictures to help you.

For the red cross, It's normal because If I want to access the vmware, I must shutdown three interfaces.

[spravtek]

Did you go through the KB I mentioned?

What does the command:

show etherchannel load-balance 

and

show etherchannel summary

Result in ?

Did you configure any overrides on the port groups?

[netgus2]

Yes, I know this KB, but for me, It's ok for my switch (but I think).

My difficulty is to see the side of what happens VMWare! How to troubleshoot this problem.

C2960S(config-if)#do show etherchannel load-balance
EtherChannel Load-Balancing Configuration:
        src-dst-ip

EtherChannel Load-Balancing Addresses Used Per-Protocol:
Non-IP: Source XOR Destination MAC address
  IPv4: Source XOR Destination IP address
  IPv6: Source XOR Destination IP address

C2960S(config-if)#do show etherc sum
Flags:  D - down        P - bundled in port-channel
        I - stand-alone s - suspended
        H - Hot-standby (LACP only)
        R - Layer3      S - Layer2
        U - in use      f - failed to allocate aggregator

        M - not in use, minimum links not met
        u - unsuitable for bundling
        w - waiting to be aggregated
        d - default port

Number of channel-groups in use: 2
Number of aggregators:           2

Group  Port-channel  Protocol    Ports
------+-------------+-----------+-----------------------------------------------
1      Po1(SU)         LACP      Gi1/0/27(P) Gi1/0/28(P)
2      Po2(SU)          -        Gi1/0/21(P) Gi1/0/22(D) Gi1/0/23(P)
                                 Gi1/0/24(P)

And for your last question, I don't know what's mean the overrides on the port groups, because I begin with VMWare vSphere. I will see on the Internet.

[lazyllama_uk]

This is a very late reply, but we just experienced exactly the same symptoms with the same switch model etc. As it happens the switch model was irrelevant.

In our case, it turned out that the management portgroup had different load-balancing settings to the vSwitch that it was on, so although the vSwitch was set to IP Hash, the portgroup was set to override that and use Port ID. No idea if that was the case for you but I thought it worth posting in case anyone else finds this question and no previous answer.