VMware Cloud Community
SHAMRIN
Contributor
Contributor

NLB issue

Hi!

We has ESX4 cluster, two VM with Windows 2008 R2 and TMG-cluster with NLB-multicast (internal and external). Through it service OWA is published and all works perfectly.

There was a problem to publish a new site, also on port 443
I have added the additional IP (second VIP)  - no alerts. Network-teams has made a static binding "new IP - new MAC" on the router. I have created the publication - no alerts and all works perfectly, but only in borders of a network of external interfaces!!!

At first I thought that it is an error of guys of network-teams.

However at check of the new publication in borders of a network of external interfaces it was found out that at test station in ARP cache to each VIP-address there corresponds the same MAC-address, namely the address primary the VIP.  Thus the new site published through second VIP, is accessible and perfectly works!

If to change in places the primary VIP and second VIP address  - second VIP correctly starts to work, and the old first  - becomes accessible only within the board of a local network.

Why?

In documents of VMware of anything in addition it is not told, only "at a multicast  mode of anything it is not necessary to adjust"

Thanks

Reply
0 Kudos
1 Reply
SHAMRIN
Contributor
Contributor

On the network equipment it is necessary to register not separate independent pars "NLB IP - NLB mac", and "primary NLB IP - primary NLB mac" and "secondary NLB IP - primary NLB mac"  😉

Reply
0 Kudos