Ping !!!!.!!!!.!!!!

[mohammad saeed]

Hi Guys,

I have two Servers C series I installed VMware Esxi v5 inside it then CUCM 8.6 and I made NIC teaming inside each server VMware and port group in my 6509 core switch which is connected directly with servers.

C220 server(VMware esxi+CUCM) -----link1-----   core 6509

                                                        -----link2------

The issue (in both servers the same) is I noticed when I ping from 6509 to IP address of CUCM the ping like this and slow !!!!.!!!!.!!!!. and when I disconnect one link and the ping become very fast and good but the issue with two cables (port group). Is it related to Jampo packet issue! or another thing, upon your experience with the same issue!

Thanks for  all

Mohammad

[Kirk J]

Greetings.

Please confirm what kind of VMware vswitch teaming type you are using (i.e. IP hash, Originating port ID, etc) and what kind of port-channel on the 6509 you have configured.

Does the output of #show int xx for those 6509 ports show any kind of error counters?

Do you have jumbo frames configured?

Thanks,

Kirk...

[mohammad saeed]

Greetings :)

I am using IP hash and I am using the default port-channel

I didn't try Jumbo frames yet!

Thanks :)

[Kirk J]

Please paste the running config for the two ports and the channel group itself#

Sample below:
interface Port-channel10
description Channel trunk to ESXi Server1
switchport
switchport trunk encapsulation dot1q
switchport trunk allowed vlan x,y,z...
switchport mode trunk
no ip address
end

interface GigabitEthernet1/10
description Esxi nic 1 
switchport
switchport trunk encapsulation dot1q
switchport trunk allowed vlan x,y,z...
switchport mode trunk
no ip address
flowcontrol send off
spanning-tree portfast
channel-group 10 mode on
end

interface GigabitEthernet1/11
description ESXi nic 2
switchport
switchport trunk encapsulation dot1q
switchport trunk allowed vlan x,y,z...
switchport mode trunk
no ip address
flowcontrol send off
spanning-tree portfast
channel-group 10 mode on
end

Confirm the hashing algorithm:

#show etherchannel load-balance

Also, it is possible to override the vswitch level teaming at the port-group level, so please confirm that your port-group/network the guestVMs are using, is also set to inherit the  IP-hash and not manually set to originating port-ID

Thanks,

Kirk...

[mohammad saeed]

here is my config:

interface GigabitEthernet1/1/38
description ## Connect to CUCM1 port0 ##
switchport
switchport trunk allowed vlan 102
switchport mode trunk
switchport nonegotiate
channel-group 32 mode on
!
!
interface GigabitEthernet1/1/40
description "Connecting to CUCM2"
switchport
switchport trunk allowed vlan 102
switchport mode trunk
switchport nonegotiate
spanning-tree portfast edge trunk
channel-group 33 mode on

-----------------------------

interface GigabitEthernet2/1/38
description ## Connect to CUCM1 port1 ##
switchport
switchport trunk allowed vlan 102
switchport mode trunk
switchport nonegotiate
channel-group 32 mode on
!
!
interface GigabitEthernet2/1/40
description "Connecting to CUCM2"
switchport
switchport trunk allowed vlan 102
switchport mode trunk
switchport nonegotiate
spanning-tree portfast edge trunk
channel-group 33 mode on

--------------------------------

interface Port-channel32
switchport
switchport trunk allowed vlan 102
switchport mode trunk
switchport nonegotiate
!
interface Port-channel33
switchport
switchport trunk allowed vlan 102
switchport mode trunk
switchport nonegotiate

---------------------------

Core_SW#show etherchannel loa
EtherChannel Load-Balancing Configuration:
src-dst-ip enhanced
mpls label-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
MPLS: Label or IP

any thing wrong?

[mohammad saeed]

I checked every things and still has the same issue! when I ping from core sw gave me: !!!!!.!!!!!.!!!!!.!!!!!

Why that ping issue happened?

Actually the CUCM works fine no issue but I am confusing about the ping result. And when I shutdown one link it took around 1 min to take over by the link2!?

Why?

Thanks

Mohammad