Use a kickstart file for red hat, no dhcp?

[Hybrid_theory]

I'm trying to do a red hat Advanced server 4 update 7 install using a kick start file on the network. The server with the kickstart file and dvd contents is on the same subnet as the server receiving the install. We have done it before in the past just fine using:

linux nofb text ks=http://serverip/path at the boot/install screen. After that the ip address info is entered manually and the kickstart does its thing. When doing this with update 7, the install acts like a regular install. When i check the http access logs on the other server, it shows no access from the server receiving the install.

I have tried adding an ip address to the install command. Such as:

linux nofb text ks=http://serverip/path ksdevice=eth0 ip=IPADDR netmask=255.255.255.0 gateway=GATEWAY IP

The results are the same. Redhat docs for AS 4 say a dhcp server must exist, but this has been done before. Any ideas?

 

[Easy Rhino]

so prior to this you did not have a DHCP server and it worked fine. this update however requires a DHCP server which you do not have and it does not work? could it be as simple as actually needing a DHCP server this time out?

 

[Hybrid_theory]

so prior to this you did not have a DHCP server and it worked fine. this update however requires a DHCP server which you do not have and it does not work? could it be as simple as actually needing a DHCP server this time out?

Well its in the Red hat advanced server 4 docs that it says this. So it shouldn't matter since it was doable before.

If it comes down to needed dhcp, we may implement it somehow. But it's a static environment, so would be good to avoid that.

 

[Easy Rhino]

so the server logs indicates that the client has not contacted it or even attempted to contact it. what does the client side say? is there an error message after you manually enter the command?

 

[Hybrid_theory]

so the server logs indicates that the client has not contacted it or even attempted to contact it. what does the client side say? is there an error message after you manually enter the command?

Nah, no error. It goes to do the install regularly. I'm using the redhat dvd to try this install. The kickstart does all kinds of automation, so even with the disk its worth it. I've setup just a bootdisk and tomorrow i will try that and hopfully force it to goto the server.

 

[Hybrid_theory]

boo urns. the server looks at the boot cd and tells me to insert a floppy. I tested the iso in vmware first and it worked fine.

 

[Easy Rhino]

boo urns. the server looks at the boot cd and tells me to insert a floppy. I tested the iso in vmware first and it worked fine.

wow that makes no sense. perhaps the update is borked?

 

[Hybrid_theory]

wow that makes no sense. perhaps the update is borked?

well ive found today through some error logs during the install, the server cannot mount the nfs file. I did a tcpdump on port 2049 on the server with the config file on it. and it received no nfs attempts. The IPs are in the same subnet, same vlan. I don't know why it wont work.

 

[Hybrid_theory]

I am such a noob. The interface on the switch was down for eth0. Weird it didn't come up in the install logs. I would have checked sooner but i didn't expect it to be administratively down.

 

[Easy Rhino]

I am such a noob. The interface on the switch was down for eth0. Weird it didn't come up in the install logs. I would have checked sooner but i didn't expect it to be administratively down.

that figures doesn't it. who messed with the interface settings on the switch? and why doesnt the kickstart app running on the client side give you atleast some sort of error message while running?

 

[Hybrid_theory]

that figures doesn't it. who messed with the interface settings on the switch? and why doesnt the kickstart app running on the client side give you atleast some sort of error message while running?

That's what I didn't get either. Instead of unable to mount nfs, say like cannot reach client at IP:addr.

Its possible the last person that had a blade in that slot was testing the backup link and shutdown the primary link on the switch side.