Is it possible to kickstart a virtual ESX 4.1 server? I'm trying to test some ks scripts in a VMware Lab and my 3.5 ESX servers are kickstarting just fine. My 4.1 servers on the other hand keep throwing nic errors.
I kickstart my ESXi 4.1 servers just fine. What kind of NIC errors are you receiving?
I don't have direct answer for you, but maybe the initial install can be done using the GUI. If the GUI finds the NIC(s), you can complete the manual install, and when complete, you'll have a ks.cfg file at /root which you can SCP out of the Host and use it as a guide.
I'm trying to configure a pxeboot / FTP setup using the Xtravirt Ultimate Deployment guide and I keep getting "no cos nics have been added by the user" errors. Sometimes the system will just hang at that point. Other times it will throw errors that the DHCP took too long to respond.
These errors also occur when attempting a GUI based install off the CD and allowing DHCP to assign an IP address. If I assign a static IP address the manual GUI install will work.
maybe this will help
Troy Clavell wrote:
maybe this will help
I've tried using the IPAPPEND 1 without any luck.
If I assign a static IP address the manual GUI install will work
Maybe it's something with the DHCP Server? Can you sucessfully run the kickstart if you specify an IP address in your ks.cfg?
Troy Clavell wrote:
Maybe it's something with the DHCP Server? Can you sucessfully run the kickstart if you specify an IP address in your ks.cfg?
I cannot get that far. If the IP address is not specifed up front the ks.cfg doesn't load because the "DHCP server takes to long to repsond" and I cannot contact my FTP site where the ks.cfg is located.
I'm wondering if this is because my ESX server is virtual. Maybe if I were using all physical servers this would work.
I'm wondering if this is because my ESX server is virtual. Maybe if I were using all physical servers this would work.
could be.. Are you using NAT for your Network Adapter? Maybe switch to Bridged.
Everything is virtual on a physical ESX server so I'm using a standard vSwitch with all VM's connected to the same vSwitch.
...i'm stuck then... If your ESX VM is on a DHCP VLAN, then in theory, it should work. But....
I've also tried manually assigning a MAC to the ESX VM and creating a DCHP reservation but still no go.
I found this blog, which may or may not be useful to you
What is the command to initiate a kickstart when booting from a CD-rom? Is it still esx ks=ftp://<path to .cfg file> ksdevice=eth0 ip=<ip_of_server> gateway=<gateway_of_server> netmask=<netmask_of_server>?
Troy Clavell wrote:
I found this blog, which may or may not be useful to you
I found that and it uses Linux as the DHCP and TFPT server as opposed to Windows. Don't want to sound biased by Windows is usually the root cause my most of my problems.
yeah, something like
ks=ftp://<server>/<path>/ks.cfg
and remember this has to come after the initial menu which will be something like
append initrd=initrd.img vmkopts=debugLogToSerial:1 mem=512M quiet
Do you have the kickstart script available to display as a post?
I'm just using the standard ks file. Nothing fancy in the pre or post sections.
I ran into this issue in the past and I ended up using the Ultimate Deployment Appliance and using the subtemplates function and hard coding the COS as VMNic0
# Network install type
network --device=vmnic0 --bootproto=static --ip=[IPADDR] --netmask=255.255.255.0 --gateway=1.1.1.1 --nameserver=1.1.1.10 --hostname=[HOSTNAME] --addvmportgroup=False
It simplified the management of building ESX servers for me. Also I tested the build within a VM and it worked fine.