cancel
Showing results for 
Search instead for 
Did you mean: 
cancel
788
Views
0
Helpful
1
Replies

Cisco UCS, ESX5.1 and NLB Cluster problem

sdawson35
Level 1
Level 1

Hello,

We have a major issue with our NLB cluster since migrating to UCS, Can anyone help at all ?

ESX5.1 host running on Cisco UCS B series blade.

2 x Windows 2008 R2 servers running NLB (unicast as multicast will not work) inside the ESX5.1 host (Distributed Switches) The servers have 2 NIC's , one nic dedicated for NLB the other for non NLB traffic but all in the same vlan The UCS blades go through 2 x 6248's running in vhost mode The 6248's are in turn multi homed to Cisco Cat 6500's The 6500's have a static arp entry for the NLB cluster address

After 4 hours the nlb cluster just dies , the only way to recover it is to move one of the servers from one vm host then back again.

Aditional information

When trying multicast the nlb cluster just dies instantly i.e : doesn’t work at all

1 Reply 1

rummanshaikh
Level 1
Level 1

Check the Mac addresss aging.

Select >> Equipment>> Policies >> Set MAC address aging to Never.

It will work.

Review Cisco Networking for a $25 gift card