Basically, I have 3 W2K3 instances and one RH9 instance running on a GSX server with a Pentium 2.4Ghz , 3GB of RAM and 120GB harddrive and some back-up drives etc.
What happens is that the Windows machines can happiy use the GSX until I power the RH9 Machine. The RH9 machine boots up (with the changed check_link_down() function added to ifcfg-eth0) and gets an I.P address. The next thing that happens is that the HOST machine requests a new I.P at the same time, and using maybe a different MAC address as all of a sudden it is using a dirty (172. - unknown host on our network ) address which has been setup so that new machines can be added. The behaviour then is that after a while, the DHCPclient on Red Hat 9 actually requests a new I.P again, and it is assigned something in the same 172. subnet. The behaviur is very odd. I have also now tested this on my Workstaion 4 software and it doesn the same thing. Add a RH host to the running systems and boom you have lost the defined I.P for both machines.
Can anyone shed any light on why Vmware would behave like this.
The odd is thing is all the Windows instances work actually continue to function.
Thanks in advance.