question

gbonasso-main avatar image
0 Votes"
gbonasso-main asked gbonasso-main edited

Microsoft Failover Cluster node not sending out Gratuitous ARP requests after a failover

DLT-DBDR01
b4:7a:f1:34:67:4b
10.10.1.16

DLT-DB02
08:f1:ea:6f:23:dd
10.10.1.12

both nodes on Windows Server 2016 last update on june 2021
We have a Windows Failover Cluster involving node DLT-DBDR01 and DLT-DB02
HKLM\SYSTEM\CurrentControlSet\Services\Tcpip\Parameters\ArpRetryCount=3

Network Interface model HPE 331li, teamed via ServerManager

situation:

servers on VLAN1001, clients on VLAN1003

we do a role (10.10.1.19) failover from node DLT-DBDR01 to DLT-DB02, and after capturing packets on the CISCO switch, we don't see any GARP request from DLT-DBDR01
131942-failover.png

so the role is unreachable from the VLAN1003

we then do a clear of the ARP table, and the role starts to be reachable.
oddly enough, if we then do a failback, the request appears on the packet capture
131953-failback.png

I thought this was a switch issue, but why the failback doesn't need an ARP table clear?

Thanks in advance,

windows-server-clustering
failover.png (17.8 KiB)
failback.png (47.7 KiB)
· 1
5 |1600 characters needed characters left characters exceeded

Up to 10 attachments (including images) can be used with a maximum of 3.0 MiB each and 30.0 MiB total.

0 Answers