2016 - Connection Broker with NLB multicast

Question

Hi,

I have a RDCB cluster with 2 servers using SQL 2016 for HA.
I have configured Network Load Balancing in multicast.

I have added the ARP entries on the cisco router, btw I get the problem localy too
arp "IP" "MAC" ARPA
mac-address-table static "MAC" "vlan id" "interface"

If I shutdown one of both server. The Connection Broker is still working
If I shutdown both of them for 10 mns. After power on, the CB is working back again
If I shutdown both of them for the night, at the morning, If I start the Connection Broker server 2. The Cluster IP dont answer to the ping
I then start the Connection Broker server 1, and now the cluster IP work.
I checked the NLB on the server 2, it only show the server 2 on the cluster. Status is converged.
I checked the NLB on the server 1, it shows both servers but server 1 is "converging" and server 2 is "converged"

Otherwise when I start NLB Manager I get a warning saying ""Running NLB Manager on a system with all networks bound to NLB might not work as expected" but my NLB is configured with Multicast, and I have 2 IPs address pro server (1 for the server and the other one for the cluster IP)

BTW, after this long offline period, after i powered on both servers, the Farm doesnt work anymore. I'm not able to connect to my collection.

I take a look to the event viewer, Terminalservices-sessionbroker and -client couldn't find something helpfull
In the overview everything seems ok

Do someone as an idea how to manage that ?

Answer 1

FYI
If I connect to NLB Manager from another W2016 server I dont get the warning, but the server is still "converging"
Both servers are hosted on VMware ESXI

Answer 2

I changed the cluster Ip Address, it's like the ARP entries bring a mess

arp [IP] [MAC] ARPA
mac-address-table static [MAC] [vlan id] [interface]

Now both servers are converged and the farm is working great.

It has helped me to write all this stuff to solve the problem :D

Answer 3

Hi ,

Welcome to our new Q&A platform.

Based on your complexity and the specific situation, if there is no specific event/error about the problem, it is hard for us to analyze from Q&A platform support level.

So after discussion with my colleagues, we suggest you open a case with Microsoft where more in-depth investigation can be done so that you would get a more satisfying explanation and solution to this issue.

You may find phone number for your region accordingly from the link below:

https://support.microsoft.com/en-us/gp/customer-service-phone-numbers

Best Regards,

Candy