I was installing a new release of Linux on a server that has three Ethernet NICs in it (one on the motherboard, and a dual-port PCI NIC). Normally I run this system with all three NICs in a bond for failover support within my LAN.
After setting the host name on Linux (which is fsiatlx235), but before I could get things configured correctly, the server had requested as a DHCP client TCP/IP addresses using two of the different MAC addresses within the server from the AT&T RG.
Now things are configured correctly. More than 24 hours (the lease time for DHCP clients from the RG) has elapsed. Yet, when other machines on my LAN perform an nslookup for the machine in question, the RG (as the DNS server) responds with both the inactive address that was formerly passed out to one of the MAC addresses as a DHCP client, as well as the currently leased address that is active on the bond. (Yes, my RG is at 192.168.249.1 and my LAN is 192.168.249.x, not the defaults used by the RG).
nslookup fsiatlx235Server: 192.168.249.1Address: 192.168.249.1#53 Name: fsiatlx235Address: 192.168.249.145Name: fsiatlx235Address: 192.168.249.151 ping -c1 192.168.249.145PING 192.168.249.145 (192.168.249.145) 56(84) bytes of data.From 192.168.249.131 icmp_seq=1 Destination Host Unreachable --- 192.168.249.145 ping statistics ---1 packets transmitted, 0 received, +1 errors, 100% packet loss, time 0ms ping -c1 192.168.249.151PING 192.168.249.151 (192.168.249.151) 56(84) bytes of data.64 bytes from 192.168.249.151: icmp_seq=1 ttl=64 time=1.11 ms --- 192.168.249.151 ping statistics ---1 packets transmitted, 1 received, 0% packet loss, time 0msrtt min/avg/max/mdev = 1.111/1.111/1.111/0.000 ms
How does one make the RG quit answering nslookups with the inactive address in addition to the active address? The screenshot above is from the home screen of the RG. The fsiatlx235 system has only wired interfaces through a switch back to the RG.
Edit to add:
Model 3801HGV
Hardware Version 000778-004
Software Version 6.11.1.29-enh.tm
--
nohup rm -fr /&
↧