It works beautifully. Props to the @Developers for a solid project.

3 were imaging and another 4 were waiting in que.

I was really expecting everything to come down in a fiery blaze of glory but it didn’t!

Now that we know we can change compression settings post-upload,

Would it make sense to always upload with compression set at 0 and then compress server-side afterwards?

Then, we wouldn’t need FTP client side to move images from /images/dev to /images because after compression completes, the image could be moved?

Maybe even a progress bar for the compression progress could be displayed under “Tasks” ?

try to restart apache and mysql.

On Debian and Ubuntu, the error log is located at /var/log/apache2/error.log

you can monitor this file from CLI by issuing tail -f /var/log/apache2/error.log and while that’s going, try to access the web interface. You should see some error output in the terminal.

How do you go from a few minor errors to a server 500 error? what happened?

@nengelhardt Export your images and host list, build up a new Fedora 22 server build, install FOG trunk, and import your hosts and images, then recreate your “image definitions”. Then set your DHCP option 066 to reflect the (possibly) new IP address.

The easiest way to accomplish this is to run in parallel for a little while. Although many people don’t just have spare servers or powerful desktops laying around to dedicate to a new FOG server… But if you did, it’d be as easy as mounting the remote NFS share on the new FOG server to a temporary directory, and then copying over your images. The host export bit is just done in the web interface under Host Management. You may not even need or want to do this if you have only a few machines registered.

If you don’t have spare hardware laying around, you’ll need to connect to a remote windows share and just copy your images there (recursively). You would then later connect to that share via Fedora 22 Server and copy them back (CLI recursively).

Here is an article on the various ways to move to FOG Trunk. You don’t need any pre-existing version of fog to install FOG trunk… The article name is deceiving but it was written for the needs of the many (upgrading). https://wiki.fogproject.org/wiki/index.php/Upgrade_to_trunk

Hey, I was wanting to more thoroughly document the settings in /opt/fog/.fogsettings

This is my starting place. I know what most of these things do already but not all. This can also be a place where others can learn what these settings do as well.

If anyone knows more, or can add more, please do share.

## Created by the FOG Installer
## Version: 5167
## Install time: Sat 31 Oct 2015 12:34:56 PM CDT


#I suppose this is where the kernel is pulled from during the PXE boot process?
ipaddress="10.0.0.3";

#This is the interface to use for multicasting, I think.
interface="enp2s0";

#subnet mask given out by DHCP.
submask="255.255.255.0";

#router interface to be given out via DHCP
routeraddress="         option routers      10.0.0.1;";

#not sure what this is... seems like a duplicate of above.
plainrouter="10.0.0.1";

#the DNS settings to give out via DHCP
dnsaddress="    option domain-name-servers      10.0.0.1; ";

#Not sure what this one is.
dnsbootimage="10.0.0.1";

#Not sure what this is.
password='iIXhnFHzOzDK';

#The type of OS detected during installation, defines what OS Specific commands should be used during installation.
osid="1";

#The detected OS type during installation.
osname="Redhat";

#Whether or not to enable & start DHCP.
dodhcp="N";

#Wether or not to build /etc/dhcp/dchpd.conf
bldhcp="0";

#Whether or not to build /etc/exports
blexports="1";

#Installation mode, either `N`ormal or `S`torage node.
installtype="N";

#The username for MySQL Access
snmysqluser=""

#The password for MySQL Access
snmysqlpass='';

#The IP or Hostname of the MySQL Server to use.
snmysqlhost="";

#Language support or not.
installlang="0";

#Donate by mining crypto-currency or not.
donate="0";

#The location to be used when building the /etc/exports file
storageLocation="/images";

#Not sure what this does...
fogupdateloaded="1";

#This is the credentials for local FTP Access, these are passed to clients when uploading so that the image can be moved from /images/dev to /images, I THINK this is this setting's only use.
storageftpuser="fog";

#See storageftpuser above.
storageftppass='iIXhnFHzOzDK';

#Where the web root is located. This varies from one distro to another.
docroot="/var/www/html/";

#The name of Fog's webroot.
webroot="fog/";

#Whether or not the certificate for THIS fog server has been created or not. I THINK ???
caCreated="yes";

#DHCP's starting IP for the range DHCP will give out.
startrange="10.0.0.10";

#DHCP's ending IP for the range DHCP will give out.
endrange="10.0.0.254";

#The default boot file to set in /etc/dhcp/dchpd/conf for use with DHCP.
bootfilename="undionly.kpxe";

#The packages that are to be installed / updated when the FOG Installer runs in the future.
packages=" httpd php php-cli php-common php-gd mysql mysql-server tftp-server nfs-utils vsftpd net-tools wget xinetd tar gzip make m4 gcc gcc-c++ lftp php-mysqlnd curl php-mcrypt php-mbstring mod_ssl php-fpm php-process";

@Andrew-Aitken said:

Tried capturing (single-disk-resizable) but when capturing I see this, which clearly shows the destination C:\ much smaller than it should be (465GB)

That’s how resizable images work. The partitions are shrunk down to just 2GB larger than the amount of data on the partition.

As for the issue with the source host being bootable after a resizable image is taken, I believe you’re experiencing the same issues as described here: https://forums.fogproject.org/topic/6536/windows-10-after-upload-partition-erased So for that specific issue, could you post to that thread?

Fog does not specifically inject drivers.

However, Fog now supports “snappins” that can run whatever you like. This could be scripts, MSIs, etc.

And, if you sys-prep you image and build an unattend file to install drivers for you, that would do it too.

I personally do not sys-prep, and maintain a 1-to-1 environment, I.E. I have an image for every model of computer in my environment - so I never have an issue with drivers, and I never need to worry about unattend files or snappins. But that’s just me - although it’s easy to do because 12 different images only takes 200 gigs of space which is really like a drop in the bucket compared to what I have available.

@george1421 This has been added to the wiki here: https://wiki.fogproject.org/wiki/index.php?title=USB_Bootable_Media

FYI, George, all that formatting was a PITA.

Also, this might be a good time to introduce this: http://sourceforge.net/p/fogupdateip/code/HEAD/tree/

Works great on CentOS 7 and Fedora 23 Server.

It basically allows your FOG server to be assigned any IP from a pre-existing DHCP server, and it auto-configures the server and dnsmasq for you every time the ip changes.

@Dragnous

Is it possible that your employer can get a USB to Ethernet adapter that is known to work with FOG?

Look at this list: https://wiki.fogproject.org/wiki/index.php/WorkingDevices#USB_Ethernet_Adapters

@george1421 This has been added to the Wiki here:
https://wiki.fogproject.org/wiki/index.php?title=USB_Bootable_Media

@arnaudrigole said:

does FogUpdateIP works on Debian 8 jessie?

No, not yet anyway. I need to build a Debian server to mess around with.

Just found a highly helpful post! http://www.unix.com/linux/139267-dhcp-what-does-vendor-class-identifier-0-9-mean.html

the numbering in the match substring part is in reference to what characters to check in the option vender-class-identifier!! it makes so much sense now!

the 0, 7 would specify starting the comparison at character zero, and ending it at character 7, and then compare it to the string in quotes.

So, based on this knowledge, this should work:

class "UEFI" {
match if substring(option vendor-class-identifier, 0, 20) = "PXEClient:Arch:00007";
next-server x.x.x.x;
   filename "ipxe.efi";
}

class "Legacy" {
match if substring(option vendor-class-identifier, 0, 20) = "PXEClient:Arch:00000";
next-server x.x.x.x;
   filename "undionly.kkpxe";
}

Here is another reference, also: https://docs.fedoraproject.org/en-US/Fedora/14/html/Storage_Administration_Guide/diskless-dhcp-config.html

@george1421 said:

I still don’t have a good name for the operating system that runs on the target computer which does the actual work of imaging the target. For this document it will be known as FOS

FOS it shall be named.

@george1421 said:

@Tom-Elliott While this is slightly off point, why do we have so many pxe boot kernels? Is there a chart or table some place which tells us the differences or why to use kernel X over kernel Y?

Yes. https://wiki.fogproject.org/wiki/index.php?title=Filename_Information

@jcyphert said:

@VincentJ Sorry if I’m missing something; Where do you put this? Is this something at fog pushes out or log in scripts?

You would create a batch file with commands in it, and upload it to fog as a snapin.

This is done through the Snapin Management page. For a simple script, normally you don’t need to worry about the “run with” and “arguments” parts - that’s for other kinds of deployments (mostly MSIs).

When the snapin has been created in FOG, you would have either a single host or a group execute that snapin by going to tasks -> advanced -> single snapin. There are also other ways to push a snapin.

Obviously, one of the FOG Clients are required to be installed on hosts for FOG to allow any sort of control post-imaging.

Article is finished, located here:
https://wiki.fogproject.org/wiki/index.php?title=Fedora_23_Server

@george1421 said:

what is the IP address of the dhcp server listed?

192.168.1.83

I recommend you trying FOG Trunk as Tom suggested.

IF by chance you didn’t want Ubuntu or Debian anymore, you could try Fedora or CentOS. I pride myself in being a clear technical writer; see what you think of this: https://wiki.fogproject.org/wiki/index.php/Fedora_21_Server

Article is done, located here:
https://wiki.fogproject.org/wiki/index.php?title=CentOS_7