Posts made by Kevin

It works using the Bypass Host Registration. We currently use that for 5 different images at my place of employment. It makes imaging a breeze, and very quick. We have our images password protected though, so only “authorized” users can image machines.

wubba,

Welcome to the forum and FOG in general. While you can use FOG like you intended. It’s not very efficient. FOG is really designed to be a business/enterprise imaging/deployment solution. So you would have 1 master image for desktops, and another for laptops, assuming you’re using XP. Or just having 1 Windows 7 image for everything. While you can make an image for every machine, it would require A LOT of hard drive space. Currently, our server has 7 images, plus bootable iso’s and i’m only using 56 GB of space on my drive. However, if you are looking at doing daily backup’s, I would recommend that you look at Clonezilla. It’s free and relatively easy to use. While FOG can do everything that you want it to, it might be a bit of a hassle and a lot of work to get it to work the way you want.

Please feel free to ask any more questions you may have, or if you have any concerns.

I agree with Leroy, just setup a FreeNAS box and create an NFS share. Should take just a couple minutes once FreeNAS is setup. FreeNAS has become one of the programs that I can’t live without. Being able to have Apple, *Nix, and Windows boxes all being able to talk with one server is invaluable, file storage, scheduled backups etc. The only downside is you need to have a beefy box for all that. If you were to use it just for FOG, you wouldn’t need much. Plus FreeNAS uses the ZFS file system which is much more robust than NTFS and even EXT4 in Linux. Data integrity is built in, much larger files can be stored (the largest single file that can be stored is 16 exabytes). It will also RAID very nicely. It’s also future proof being a 128 bit filesystem.

So i’ve rambled on a bit about FreeNAS. But hey, it’s Free, takes less than 10 minutes to install from scratch, proven reliable, and everything can be managed through a web interface. It doesn’t get much easier than that.

Interesting, never knew that existed Thanks

Personally, I would like to see what jdd49 changed in the .29 bypass file and replicate it once .33 comes out. AFAIK the new version of FOG used Part Image? It would be nice to run the bypass with the new “imaging software”.

astrouga, what are you using the modify the file. I tend to use gedit to modify my files but init is to big to open. Are you using something else like Vi?

With FOG .32 if you were to do a quick image, it forces a check disk on restart, regardless of how the image was setup. If you were to go through and setup the machine to deploy an image from the webserver, it’ll run just fine without running check disk. The file that jdd49 made, the bypass, would be similar to doing a quick image. It does work really well, and has saved us a lot of time doing mass imaging, but due to the mass imaging, that’s why I was asked if the check disk could be removed. And quite honestly, I have no idea where to start to even look into modifying the .32 (or ,33) init file.

Have you tried updating the kernel?

Awesome, thanks much.

Falko, would you happen to have that guide still from your DB? I had my computer crash and lost all my guides/notes I had made and found. Just trying to rebuild the collection. Being a fully encrypted RAID 0+1 setup, there wasn’t much chance of retrieving the data Thankfully, the VM’s are on a separate drive so I still have those.

Basically, by using the bypass, it’s essentially doing a quick image and if I remember right, a quick image forces a checkdisk. I don’t have a problem with it, but my job is not to replace/install computers. When we receive computers in, it’s not unheard of to have 500-800 come in at one time but can only image about 20-30 at a time. Needless to say this takes awhile and the less time it takes, the better. Hence the request made by co-workers.

My co-workers have been asking if there is a way to disable checkdisk after a machine is imaged. It doesn’t bother me at all, but it is a concern to them. We run the bypass which shouldn’t be a big concern, but wasn’t sure if there was any config file we could modify to make them happy.

Just a thought, if you create a group in FOG and move all the PC’s you want into it. Then you could make a batch file with “shutdown /s /f /t 0” and save it as shutdown.cmd. Then you can upload that file as a snap-in to fog. Lastly, if you shut the computers down at a certain time at night everyday, you should be able to set the schedule the file to run using the Cron execution through FOG. Just a thought, i’ve never tried it, so I don’t know the results.

while i’m thinking about it, is there a way to find your config during the initial setup of FOG? It sure would be trying to track down all the IP’s used for setting it up if you can just pull up your initial config.

Well, we don’t use FOG for much besides imaging. I really am the only one to administer it at this point, and keep the server up to date. So we really only use it to image machines using the bypass, and to boot recovery tools and various iso files. However, we are going to start deploying laptops with SSD’s installed. So it’s possible that that 64 could make a difference, I assume, when we increase write speeds with the SSD, but it’s still an unknown. Our server is very high-end, even though FOG is running VMware ESX. We are just trying to achieve the highest speeds possible due to the volume of computers we need to setup with an initial image, or re-image. We’ve discussed trunking ports on the server to try and double the bandwidth to FOG. As it stands, we can image a machine with a 10 GB image in 3 minutes 45 seconds. The problem comes when we need to image 500+ computers. Multicasting proves to be to slow in the terms of hours to image the machines, so we stick with unicasting. But I think I might just have to setup a test server on 64bit for a while and see what kinds of results I can get.

So, now that Ubuntu 12.04 is out, and has been for a couple months, i’m starting to look at making the jump to 12.04. Now the fun question, is there really any advantage to using 64 over 32 bit Ubuntu? My test box I have setup for FOG is running 10.04 LTS 64 bit, however, our production server (which has quite a bit more power) is only running Ubuntu 10.04 LTS 32 bit. Does anyone prefer either architecture to another? If we switch to a 64 bit OS, will there be any noticeable difference to imaging either to upload or download?

It wasn’t up to me unfortunately. I was told that we can’t use the client based on the fact that it would cause to much “traffic” on the network with it checking in. Our way of imaging computers was very backwards. A group of us built FOG in an office to get bugs worked out until we were ready to demo it. The boss liked what he saw, the fact that it was much simpler to setup machines, and the fact that you wouldn’t have to remember cd’s or flash drives everywhere you went, we could PXE boot from any machine. So we were asked to put our server in the DC. When we brought up installing the client on the machines, we had a test pilot group of computers which we were very successful with. However, we were told that we were unsure of the traffic it would generate being on every machine. So we are forced to look at doing the images without the clients installed. BTW, I’m not in charge of our images, I was just sick of all the little steps we had to take to get an image installed on a computer, hence why we put FOG together

The short answer is yes. I know this because we currently have FOG running with WDS. However, we are now looking at going with just FOG for everything. We currently run a *nix DHCP server and run FOG off of that. We then have WDS setup on another server and boot an iso to access that server. So we don’t have DHCP turned on for WDS, but the iso points directly to the server itself. So we can use FOG to boot the iso and get us out on the WDS server ;). But WDS seems to be more hassle than it’s worth, so after working with falko from here, finally getting a working Windows 7 image that can be pushed through FOG just like the XP images. Then I have the bypass host installed so I can literally choose windows 7 from the PXE menu and install it on any machine on the fly.

The guide worked perfectly. I just made some minor changes and got it configured the way I wanted to, so it’ll only prompt me for a PC name. However, the last hurdle, and apparently a well known issue, auto joining to the domain. From the way the answer file works, it’ll join the PC to the domain prior to being named, giving it a random name in AD. I’ve found the following command that can be used but have a question on one piece of it.

[B][I]NETDOM /Domain:YOURDOMAIN /user:DOMAINADMIN /password:DOMAINPASSWORD MEMBER [COLOR=#0000ff]COMPUTERNAME[/COLOR] /JOINDOMAIN[/I][/B]

I know that you have to have the AD tools installed to used this, as netdom is not part of windows 7 by default. However, to truly automate the process can the computer name in blue be set to something like %pcname% to automatically pull in the name of the computer? With us having close to 10,000 machines, we would like to have as much of the imaging process automated as possible. I’ve seen a powershell script and also a vbs script that claim to do the same thing, so i’m curious to see if the above post would work, or if there is an easier way yet to accomplish this task.[B][I] [/I][I][/I][/B]

Did you do just do a quick host registration? If so, the laptop won’t have an image or an OS ID associated with it.

So I tried setting this up last night, and was going good until I tried loading a sysprepped image. I did a “hybrid” approach to this. As of right now, I’m trying to get this to work without the FOG client, as we do not use the client at my place of employment. I’ve setup a generic image in a VM, install the programs, ran my updates etc. I then used this document and followed the instructions to make the autounattend file, [url]http://www.bauer-power.net/2011/02/how-to-sysprep-windows-7-on-easy.html[/url]. However, when the sysprepped image loads, I get a “Windows could not parse or process the (generalize)…” So i’ve been trying to figure out what has gone wrong. I think i’m going to just bite the bullet and create the same file in WAIK. So does everything that needs to be setup in WAIK need to be done under the specialize field? Or did you pretty much just do this portion from the guide you listed (of course, adjust as needed):

MY IMAGE IS x64 SO I ADDED THE AMD64 OPTIONS

1. Set Owner Information

2. Enable the administrator account and set password

3. Skip Activation

4. Hide License Info

5. Protect PC = 1 (basic protection)

6. Network Location = Work

7. Hide Wireless Setup

8. Set Language to UK

9. Show Windows Live = false

10. Skip Rearm = 1
    This allows you to run sysprep more than 3 times on a machine, but i would recommend NOT using it.

[url]http://technet.microsoft.com/en-us/library/ff716063(WS.10).aspx[/url]

11. copyprofile=true
    copy profile will copy the administrator profile on sysprep and make it the default for any new user login. could be handy (i will use)