It seems to me that you need to tunnel your devices. If you have more than one device I would use that to create a tunnel (point to point VPN) to the other side and then go through that. In other words:

PXE <—> DHCP Server w/VPN to Other network and routing <----we shouldn’t care what is here----> Other side of VPN <—> TFTP server

I don’t know why that wouldn’t work. Now doing an “on demand dial vpn”… that I’m not sure of.

On the DHCP server you should have a static route that would go through it to your TFTP server. My understanding as to how it works is that you send a special DHCP request (your PXE boot) to the DHCP server which says “here is an IP now you want to go here” (that is the helper IP for the special PXE request). From there it is only a matter as to HOW to get there. If you have a VPN tunnel established then you should be able to route across that and hit the TFTP server.

Your only issue at that point is more TFTP than anything. I don’t know how it works with FOG but when the TFTP image is brought down if you have issues on your network it will fail and have to start again. What I do not know is how FOG sends the image. if it is TFTP then you have to keep your fingers crossed. If it is more involved and accounts for network congestion/issues during the transfer then it should be fine.

Wow I would never imagine that anyone would actually read this after so long.

I can say that since there was such a severe lack of an update/cahange in so long nor any word from the devs either my company and I realized it was time to look for an alternative. We ended up going with the new Windows Deployment Services and Windows Deployment Toolkit.
Let me first say that it is NOT the same HOWEVER… it is such a breath of fresh air that we don’t even care.

You can do 9 above which makes things so awesome. You can essentially create a deployment where you become like an OEM install. You can still perform the same steps to make ‘snap-ins’ for software you want to install that is not msi format.

Essentially instead of needing an “image” of a machine you essentially build the image on the fly. So far our testing shows that it’s about the same time elapsed to do it either way. The reason is that the WDS doesn’t really need an image because you feed it drivers for a machine and part of the boot checks the hardware and matches the drivers up and goes. That in conjunction with the new USMT tool it really makes reimaging a breeze.

Our next step is to use another tool that I cannot remember the name for the life of me to manage updates and software version etc. For the user login stuff we have moved to Spiceworks which tracks that.

Ok so yes I am fairly new to FOG and maybe some of this can be handled in a different manner. I am also leaving out things others have said; like single snap-in to multiple machines.

1. it would be nice to get more information on screens like the Hosts screen. Like for example have the Fog Service update the server with the IP address and perhaps the last time it was contacted etc.

2. Following 1… have a column for Group on the hosts screen

3. Have a column for “Date Added/Imaged/Last Imaged”

4. Have a column for “Last Logged in User”
   Basically have more information available on the host screen that which by default should SHOW ALL HOSTS and become like a “control room” kind of thing.

5. Be able to see “Group” relationship once you are inside a host. Right now I have to pull up each group then look at “membership” and then look for a machine to see why it may be having issues. I should also be able to open a host, verify that is the host in question and then from that screen add it to the host. Currently from what I can tell you have to add group memberships on the “Show all Hosts” or the “Search Hosts” results.

6. Further expanding… you should be able to pull up a host and see what settings were placed from the “GROUP” as opposed to directly on the host.

7. Please update the Fog Service/Fog Client. These need serious overhaul to get rid of the issues I’ve seen around here (the forums). I myself am having issues with Win7 machines and printer deployment/snap-in deployment and “Interactive Services” pop-ups and interruptions to the user.

8. Mask the password in the “Add to AD after imaging screen”

9. For a deployment there should be a deployment flow screen that you can “inject” pieces into. For example you can say 1) Run X Script, 2) Join AD, 3) Run Y script, 4) install snap-in X etc. that way you can really tweak the process of deployment to hopefully get it 150% hands off in the process.

Thanks for listening.

If anyone sees something here I am wrong about or different way to achieve them already then please let me know.

So I have a question… it has worked in the past with XP however with Win7 it doesn’t.

We purchase HP machines from our vendor and so each has it’s own Win7 Key. However since it is an HP they have their “special OEM key”. With XP we were able to rip that serial using jellybean or whatever program and then put that key in our sysprep and everything would work as advertised.
With Win7 however it will not accept that key. I am assuming we may be able to still do it if we started with the installed OS (the one with the bloatware) however when we tried that in the past it adds ~40 minutes to the image upload/download (for some strange reason).
Anyone know of a work around for this?? I may just be going about this the wrong way.

Also with Windows 7 and doing snap-ins and printer management on Win7 boxes whenever we push a snap-in or a printer it doesn’t install and instead invokes the “interactive services” screen which then normally comes up with an error etc. It works fine on XP, only does this in Win7. From what I’ve understood this is because of the way User0 is not used anymore in Vista (we have none of those) and Win7 so if the program isn’t updated it will not work properly. It doesn’t always push printers and/or software and now constantly keeps bugging users with the interactive services pop-up and I cannot find a way to get it to stop.
Anyone have any ideas there?

I am obtaining a 2nd HDD so I can just duplicate it as well before performing update. Thanks.

I’m having a similar issue where I can register fine with full host registration however when we install the client on a PC it will not register the host with the client. I got an error of “Only 1 MAC address”. Any ideas?

Good Afternoon,

I currently am running .29 and am looking at updating to .32. First thing I want to ask is there any document (I have not been able to find any) that describes this process? I have found that the wiki said that updating was as easy as rerunning the installer and it will update. What I am looking for moreso is can I go directly to .32 or do I need to go .30, .31, and then .32?

Lastly how can I/what do I need to do to back up my system in case I need to recover??

How about this because I am looking for this also and apparently this has been answered…

Can you make a section here to request a thread from the old forum to be brought over to this one? At least then it’s not all but ones with hopefully useful information. The one I am looking for has to do with the Fog Client Service not registering hosts and the link I found doesn’t have a google cache. [url]https://sourceforge.net/projects/freeghost/forums/forum/716419/topic/3727752[/url]

I got an error on my fog.log saying that the host registration exited because there was only 1 MAC address if anyone know about that in particular.

@Bruce
I do not think you will be able to do what you are thinking with the Snap-Ins. From what I gather it’s not like an “automated task” but more so an app deployment tool that is easier than using say AD to push software/settings etc… The way I’ve used it is get “WinInstall LE” which basically takes a snapshot of your computer and then says “go”. You then do whatever you want to the PC (like install an application) and then tell it “I’m Done”. It then takes another snapshot of your disk and records the changes; like a new directory with files in it and registry keys etc. Then it takes all of those changes and packages them together in one file; your snap-in.

Then you just push that and it puts all those changes on whatever computer you have told it to be on.

The reason why I say it’s not like a tool where you can do administrative things (like say run tools and some things) is that you don’t feed it “live information”. We have basically kept it to one program and/or setting change per snap-in so we don’t get confused or try to do too much and have to change too much when something is updated etc.

Then I would assume it should be ok once you have gotten to the part where the machine is pulling the Image down.

It seems to me that you need to tunnel your devices. If you have more than one device I would use that to create a tunnel (point to point VPN) to the other side and then go through that. In other words:

PXE <—> DHCP Server w/VPN to Other network and routing <----we shouldn’t care what is here----> Other side of VPN <—> TFTP server

I don’t know why that wouldn’t work. Now doing an “on demand dial vpn”… that I’m not sure of.

On the DHCP server you should have a static route that would go through it to your TFTP server. My understanding as to how it works is that you send a special DHCP request (your PXE boot) to the DHCP server which says “here is an IP now you want to go here” (that is the helper IP for the special PXE request). From there it is only a matter as to HOW to get there. If you have a VPN tunnel established then you should be able to route across that and hit the TFTP server.

Your only issue at that point is more TFTP than anything. I don’t know how it works with FOG but when the TFTP image is brought down if you have issues on your network it will fail and have to start again. What I do not know is how FOG sends the image. if it is TFTP then you have to keep your fingers crossed. If it is more involved and accounts for network congestion/issues during the transfer then it should be fine.