ESX deployment appliance v1.05

[brugh]

https://solutionexchange.vmware.com/store/products/esx-deployment-appliance-eda

ESX Deployment Appliance (EDA) is a small and easy to use appliance

that makes deploying ESX servers a breeze. It has a very intuitive

web-interface that can configure and deploy dozens of ESX servers in

minutes. It has a script-builder that will allow any admin to create

%post-scripts that will do most anything one needs to get the ESX hosts

up and running! Even if deploying with RDP/Altiris or the UDA, this

script-builder can help setting those up very quickly.

[brugh]

it will run perfectly from esx

what's probably missing is the next-server entry from your dhcp scope. check if it is the appliance's IP address. that's where it will look for the tftp server.

[marklemon]

, so sorry, my humblest apologies. I have been looking at the left column date which shows when a user joined these forums, not the date I should have been looking at which actually shows the date for each new post.

Again, apologies!

I'll get the original --ondisk bit and check to see what I'm missing. Anyone on HP (cciss/c0d0) have an example of their working partition bit of the anaconda.ks file I can try?

cheers,

Kevin

[marklemon]

Brugh, you are THE MAN! I don't know what was wrong with my original diskpart commands, as I've changed them at last a dozen times, but by cutting/pasting your example from your original applicance, it worked for me like a charm! Now I can do the fun bit of customizing the network and tweak the post-exec stuff. Great fun!

(by the way, it looks like only the original poster in this thread can give points (like HP's ITRC), is that correct? How can I give kudos for a spot-on answer to my question here?

cheers again Brugh, and if anyone else needs reminding of a working ks.cfg disk parittioning section for a HP DL series server:

clearpart --all --drives=cciss/c0d0 --initlabel

part /boot --fstype ext3 --size 200 --ondisk=cciss/c0d0

part swap --size 544 --ondisk=cciss/c0d0

part / --fstype ext3 --size 5120 --ondisk=cciss/c0d0

part /var --fstype ext3 --size 4096 --ondisk=cciss/c0d0

part /tmp --fstype ext3 --size 2048 --ondisk=cciss/c0d0

part None --fstype vmkcore --size 100 --ondisk=cciss/c0d0

part None --fstype vmfs3 --size 1 --grow --ondisk=cciss/c0d0

(takens straight from EDA)

[bggb29]

That was my problem next server was not modified.

The next issue I ran into was the install changing the nic so that when the network restarted

it was not pulling the ks config and I was getting the language question.

The last problem I am running into is it does not appear to set the root password to what

I think I am setting it to.

thanks

[brugh]

what root password are you trying to change? try setting it to something trivial, see if that works.

also, some special characters like & and > don't work well with webpages. i should perhaps do some more escaping in there but for now, try to avoid those.

[brugh]

i think you're right about the points. doesn't matter though; i'm not here for the points. if it works for you than that's good enough for me

[bggb29]

it was special characters, for the esx servers password. simple text is fine

thanks

works great

[BigIUFan]

I'm trying to determine if I can accomplish the following.

Is there a way for me to use the EDA on a subnet that has an existing DHCP server AND serves PXEBOOT images via Altiris?

I'm trying to configure it so I don't step on any toes while allowing me to utilize the EDA to deploy ESX in our environment.

Thanks for your time.

[brugh]

that's actually pretty easy to do.

first disable the dhcp server of the EDA

then add reservations to your current dhcp server by adding option 66 and 67 (windows dhcp) or next-server and filename (linux dhcp) for the MAC addresses of your ESX servers' first NIC.

and that's it when the server boots, it gets pointed to the eda and will pick up the bootfile from there.

[marklemon]

EDA really does Rock. I just watched a video of a presentation at VMWorld last month where this was demo'ed. Great that you are getting that kind of exposure and that VMware is recognizing this appliance as the premier way to automate ESX server rollout.

My script is working great, applying ESX patches, configuring my network and virutal network, etc.

One problem I have is that I am trying to modify my LVM and Disk settings undre “Advanced”

settings in vCenter.

I see that /proc/vmware/config/Disk/SANDevicesWithAPFailover is

where I query my current setting (5th field), but how do I change it?

What vmware-vim-cmd or vimsh parameters would let me change the value for this?

I’ve read and searched xtravirt’s great white paper, but have not been able to find or guess the syntax I need.

Similarly, I want to change “LVMDisallowSnapshotlun”. under LVM advanced settings, which can be checked with

"cat /proc/vmware/config/LVM/DisallowSnapshotLun" and see the 3rd field is either "1" or "0". But how can I change it from the service console/script?

cheers,

Kevin

Update: ah, Got it! There's nothing like trial and error:

1. vmware-vim-cmd hostsvc/advopt/update Disk.SANDevicesWithAPFailover string "SANsymphony :"

Message was edited by: KevinC

[Vampire]

Hi, I'm just testing the EDA and found one bug when I setup a new host. I want to name the new host "viesxeedv01", but when I save my host settings the "0" got lost and the hostname is saved as "viesxeedv1". Also the "ks" config file shows the hostname without the "0".

As enhancement it would be great to choose the keyboard layout during konfiguration so that the german keyboard layout is automatically choosen during setup.

Is there any chance to edit the "ks.cfg" file directly on the EDA server to change some things after configuration via web, which are at the moment not possible, like the keyboard layout?

Regards

Frank

[brugh]

if you enter the hostname with ## it will fill in the numbers with leading zeros.

and at the moment there's no way to configure the keyboard layout from the web interface during installation. you could edit the /var/www/ks/ks.php file which contains the basic ks config file and change it there. or you can put a script line in the %post to change it (put 'loadkeys' in your /etc/profile or edit '/etc/sysconfig/keyboard' using 'sed').

[brugh]

glad you like it i think it really is a very cool appliance

as for vimsh, it's very powerful and very poorly documented. good you found it though!

advanced settings for ESX can easily be changed with 'esxcfg-advcfg -s 1 /LVM/DisallowSnapshotLun'

[admin]

Hey Brugh

Almost have a great solution with your appliance (I had some switch problems caused by stp)

I only use software initiators and iSCSI arrays, so no danger of wiping LUNs on boot. I would like to keep my storage drivers in place. I have followed your advise about commenting out line 53 in /etc/apache2/sites-enable but it seems to make no difference. Any other things I could check?

[Vampire]

Thankx for your prompt answer. As you wrote using "##" to get leading zeros worked for me. I have edited the ks.php to change the keyboard layout during installation.

[brugh]

well, the 'import pxe bootfiles' actually does a couple of things. it copies the kernel and initrd files to the tftpboot directory and it strips the lan drivers from the initrd and the stage2 file (netstg2).

the change in the apache config makes it so the appliance just offers the original netstg file from the CD instead of the changed one it created during pxe import. but the initrd file is still changed and offered to the booting system. normally this shouldn't be a problem since it uses the drivers from netstg to initiate the hardware before starting the partitioning. that means if it gets the original netstg, it should be able to install for a boot from SAN setup.

i'm not exactly sure what the issue is you're still having but if you need SAN drivers in the initrd, you could import the pxe bootfiles, then go into the appliance and copy the initrd from the cd to the tftpboot directory. the original initrd is /var/www/esx/images/pxeboot/initrd.img and the place to put it is /var/lib/tftpboot/. once that's done, eda should offer only original files to your installing system as if installing straight from cd. if that doesn't help can you describe what your issue is?

[admin]

Thanks Brugh, I will give this a try. I just need the original software iSCSI device back, vmhba35.

[admin]

Ok, I tracked my missing iscsi adapter down to an unrelated problem. Your appliance is working perfectly now! :smileygrin:

[brugh]

cool another happy customer!

if i had a dollar for every hour saved .. or for every server installed... :smileylaugh:

[virtualizeme215]

Hello Brugh,

Like the others. I absolutely love your appliance. It has saved me countless hours.

But I was looking around today through my Infrastructure and I noticed something weird. Host's built using your EDA appliance is showing 250mb less memory then systems built without it......

They are identical Dell PE 2950's with 8x8GB Dimms. A system built with your appliance shows 62.75GB, and a system built without your appliance is showing 63.00GB for memory.

Do you know why this is happening?

Thanks in Advance