I am trying to get several ip addresses to be attached to a virtual server running ubuntu 10.04 on vmware esxi.
using esxi i have assigned it 10 virtual nic's which are all connected to the same virtual switch.
each virtual nic has its own ip address which no other device on the network is using.
The problem is that from within the office, i.e. before the router everything works just fine no problems, all services can be used just fine. from outside the office, i.e. public internet only one ip on the machine provides any services. that is the ip on eth0. all the other ip's time out.
all ip's are public ip's in the same range. There is no firewall issue because the firewall configuration has not been changed, and all ip's were publicly available on the old hardware that has been powered off.
traceroute outputs from another server elsewhere are as follows:
root@jaguar:~# tcptraceroute x.x.x.252
traceroute to x.x.x.252 (x.x.x.252), 30 hops max, 60 byte packets
1 * * *
2 * * *
3 Xe0-1-1-0-grtontl1.red.telefonica-wholesale.net (213.140.51.181) 21.845 ms 22.682 ms 23.545 ms
4 GE6-1-0-0-grtlontl3.red.telefonica-wholesale.net (84.16.12.14) 24.144 ms 24.834 ms GE7-0-0-0-grtlontl3.red.telefonica-wholesale.net (84.16.12.94) 25.374 ms
5 CW-2-1-0-0-grtlontl3.red.telefonica-wholesale.net (213.140.52.62) 67.339 ms 67.334 ms 67.327 ms
6 xe-4-2-0-xcr1.lnd.cw.net (195.2.25.58) 86.102 ms 64.355 ms 64.136 ms
7 ge-5-0-0-dcr1.lnd.cw.net (195.2.25.2) 18.486 ms ge-1-0-0-dcr1.lnd.cw.net (195.2.25.122) 18.664 ms 18.613 ms
8 so-0-0-0-zcr1.lnt.cw.net (166.63.163.222) 18.830 ms so-4-0-0-zcr1.lnt.cw.net (166.63.161.146) 18.437 ms so-0-0-0-zcr1.lnt.cw.net (166.63.163.222) 18.372 ms
9 so-1-3-0-ycr1-lo7.cw.net (166.63.222.30) 22.153 ms 22.413 ms 22.411 ms
10 ge-0-0-0-kar1.lo7.cw.net (195.2.12.178) 22.886 ms 22.404 ms 22.356 ms
11 hackney-gw.uk.insnet.net (213.38.241.146) 28.065 ms 28.818 ms 29.410 ms
12 * * *
13 * * *
14 * * *
15 * * *
16 * * *
17 * * *
18 * * *
19 * * *
20 * * *
21 * * *
22 * * *
23 * * *
24 * * *
25 * * *
26 * * *
27 * * *
28 * * *
29 * * *
30 * * *
root@jaguar:~# tcptraceroute x.x.x.244
traceroute to x.x.x.244 (x.x.x.244), 30 hops max, 60 byte packets
1 * * *
2 * * *
3 * * *
4 GE7-0-0-0-grtlontl3.red.telefonica-wholesale.net (84.16.12.94) 23.450 ms 24.431 ms GE6-1-0-0-grtlontl3.red.telefonica-wholesale.net (84.16.12.14) 24.956 ms
5 CW-2-1-0-0-grtlontl3.red.telefonica-wholesale.net (213.140.52.62) 25.626 ms 26.377 ms 26.975 ms
6 xe-4-0-0-xcr1.lnd.cw.net (195.2.25.30) 27.823 ms xe-4-2-0-xcr1.lnd.cw.net (195.2.25.58) 18.397 ms 18.448 ms
7 ge-5-0-0-dcr1.lnd.cw.net (195.2.25.2) 18.660 ms ge-1-0-0-dcr1.lnd.cw.net (195.2.25.122) 22.714 ms 21.367 ms
8 so-4-0-0-zcr1.lnt.cw.net (166.63.161.146) 18.789 ms so-0-0-0-zcr1.lnt.cw.net (166.63.163.222) 18.915 ms so-4-0-0-zcr1.lnt.cw.net (166.63.161.146) 18.532 ms
9 so-1-3-0-ycr1-lo7.cw.net (166.63.222.30) 22.877 ms 22.694 ms 22.405 ms
10 ge-12-0-0-kar1.lo7.cw.net (195.2.12.182) 22.404 ms 22.430 ms 22.600 ms
11 hackney-gw.uk.insnet.net (213.38.241.146) 27.783 ms 27.810 ms 28.225 ms
12 rsvd-company-name-244.x.x.x.in-addr.arpa (x.x.x.244) 28.292 ms 28.295 ms 27.762 ms
the address ending in 252 is one that does not work
the address ending in 244 is one that does work
/etc/network/interfaces looks like:
auto lo
iface lo inet loopback
auto eth0
iface eth0 inet static
address x.x.x.244
netmask 255.255.255.0
broadcast x.x.x.255
gateway x.x.x.x
dns-nameservers x.x.x.x
auto eth1
iface eth1 inet static
address x.x.x.252
netmask 255.255.255.0
gateway x.x.x.x
broadcast x.x.x.255
...
etc etc
Best Answer
Finally found the solution.
It was not anything to do with the configuration on Ubuntu or VMWare but was in fact the router itself. It's arp table seems to have an extremely long lifetime. Power cycling the router solved the problem, and has solved the problem every time it has occured since then.
Just posting this so that anyone else who might find this situation may find some help.