This content has been marked as final. Show 6 replies
i want to share one more thing. Both nodes have each NIC connected with the LAN and the ip series on both NICs are 10.x.x.x.
I saw the arp cache and found two LAN/Network computers was have the MAC addresses in arp table which were trying to acquire the IPs of NICs which are connected cross.
I dont know how the cross cable connected NICs can be conflict by the Network computers IP
Edited by: ZahidHaseeb on Aug 25, 2010 7:38 AM
The error message seems pretty self explanatory.
You have two computers both trying to use IP address 192.168.0.1
So you've probably made a mistake on the computer you list as being 18.104.22.168 and have it set for 192.168.0.1 instead.
Check the mac address of the two computers and see which one has the MAC address list in the error.
Ok, my mistake.
So that PC is also trying to use 192.168.0.1.
And it must be doing arp broadcasts on the LAN.
Solaris should probably be ignoring them since it has an 10.x.x.x address on the lan. So it shouldnt care about 192.168.0.x addresses on the LAN.
Only on the crossover network.
But apparently it cares anyway.
So I suggest you change the cluster heartbeat addresses to something else thats less likely to be chosen by random computers on your lan.
Or figure out why the PC is trying to use 192.168.0.1. It could be someone running vmware or virtual box.
Or vlan you lan, so your critical servers arent sharing a network with random poorly configured PC's.