VMware Cloud Community
AFRUM
Contributor
Contributor

Automated deployment through EDA

Hi all,

First post, and I'm not real comfortable with what I'm working with, so bare with me here. I am trying to deploy ESX 4.0 Update 1 to several Dell 2850s, 1950s, and 2950s using the ESX Deployment Appliance (EDA). I've configured EDA to be on our vSphere VLAN (14) while our NFS servers are on a storage VLAN (18). I intend for the Service Console to be on VLAN 14 and for the VMkernel to be on VLAN 18 on a separate vSwitch.

PXE boot seems to work as I get to choose which host I'm installing, and then it begins to boot. However, it eventually says "No COS NICs chosen by user" and then asks me to press a key to reboot. If I look at the esx_install.log I see that it's failing to pull the kickstart file with an error saying that it can't download via HTTP claiming there's no route to host. However, if I try to use wget to download the file from a shell, I'm able to get it and read it without an issue. I'm not sure what I'm doing wrong here, and none of the posts I've found so far seem to have any good info regarding these errors.

What am I doing wrong here? I've confirmed that vmnic0 is assigned to the "download Portgroup" that I see on the vswitch, and that NIC is connected to a port on the switch that is configured to access VLAN 14 (where the EDA is connected). I just don't see how this could be a networking issue.

Any ideas? Thanks in advance!

0 Kudos
2 Replies
shuguet
Enthusiast
Enthusiast

Hello,

I'm having the exact same problem here (EDA 0.90 + ESX 4.0 U1 => No route to host).

But to make the problem worst, my EDA and my ESX are on the same subnet, there should be no routing at all in my situation ...

Did you find a solution to your problem ?

Did anyone ever faced this problem and managed to solve it ?

Thanks in advance...

Sylvain Huguet vExpert 2014, 2013, 2012 & 2011 VCP4&5/VTSP4/VSP4&5 Nutanix NPP/NPSE/NPSR
0 Kudos
AFRUM
Contributor
Contributor

We were able to resolve this issue by modifying the /var/lib/tftpboot/pxelinux.cfg/default file and changing the "ipappend 1" lines to "ipappend 2". This, from what I've read, tells the machine to use the IP assigned by the initial DHCP lease to pull the kickstart file instead of using another IP. Whatever it does, it worked. Smiley Wink

Note that every time you save a change in the EDA web interface, the default file is rewritten, so all the ipappend values are set back to 1. This also appears to occur on a reboot.

0 Kudos