Since this is general chat and there is a lack of thanks, then a big set of thank you to the devs and helpers on here, in particular Tom Elliot, CML and Wayne Workman.

I have just finished multicasting 2 rooms of 25 PCs each simultaneously with another 50 ongoing now and another 100 to do this afternoon and a final 100 on Monday (these are harder as they have supplementary software to install afterwards). No issues and no stress (3GbMin multicast speed for those who are interested switches are cascaded netgear GS748TS, I can unicast at 9GbMin). Culmination of a months worth of on-off upgrading!

All automated from PXE first boot with fog sending the WOL too. Image goes straight to CTRL-ALT-DEL after image, OOBE, driver install, rename, join domain. No keypresses whatsoever.

FOG trunk 5370 (not the latest I know)
Centos 6.7 guest on a clustered 2k12R2 host. Guest has 2gb Ram, 2 cores assigned (htop shows the cores are not stressed)
Image is a windows 8.1 sysprepped with snappydriver driver script on OOBE - 35gb real size.
Golden image created as a VM
Network core cisco 3750 cascaded into netgear 748TS into 724TS. Jumbo frames enabled, QOS disabled temporarily to allow full multicast (to hell with normal traffic!).
(edit - storage for images is from a synology RS815+ running in pure iscsi block mode, vm host has a VHDX saved on it. Drives are standard sata 7200 3Tb drives in Raid6 so not the fastest but good enough to use Gb lan. Connection is multilink but for redundancy not speed - not that a single image pull would use MPIO anyway)

25minutes from switch on to usable CTRL-ALT-DEL

old topic but I had a similar issue, I simply created a DHCP reservation for the problem laptops (I only had a few) and set the 67 for those laptops on the reservation. That way I could leave fog alone and let the DHCP service give the appropriate details out (and being a reservation it is a permanent solution as far as I care).

Fixed.
Used https://wiki.fogproject.org/wiki/index.php/Modifying_the_Init_Image and modified /bin/fog.download

#fix for possible dodgy partitions on the disk
testdisk="";
testdisk=`fogpartinfo --list-devices 2>/dev/null`;
#this will either give us something or nothing
if [ -z "$testdisk" ]; then
  echo;
  read -r -p "Partition error. Do you want to murder the partition table for $hd? [y/N] " response
  if [[ $response =~ ^([yY][eE][sS]|[yY])$ ]]
  then
    #murder the partition table - this shouldnt need a restart
    dd if=/dev/zero of=$hd bs=512 count=1 conv=notrunc;
    #label the emptiness - should be enough to get the system to continue
    parted $hd --script -- mklabel msdos;
  fi
else
 echo "checked $testdisk for partitions (no need to kill partition table yet)";
fi

I put this before the imgtype checks

# Generates the partitions.  If singledisk resizeable, it generates
# based on default normal values.
# Otherwise it uses MBR/GPT to generate the partition table(s).

After that I get a VERY healthy speed increase (PIGZ 7 uploaded on a 4core 4gb @ about 1gb/min). This client is a modest 3ghz Core2Duo with 2gb RAM.

Testing was done on multiple brands of HDD and 2 brands of SSD reproducible. If the drive had already been cloned (some had been cloned from previous upgrades) then the system worked without my fudge. If the drive was brand spanking new out of the packet then there was a chance this would happen (both sandisk and Kingston SSDs were totally blank, the Seagate was blank but the WD was not and worked without the fudge)

@Wayne-Workman yes. Not sure if it is the cleanest of solutions though, however it will work for me, if it gets to this point in the cloning process then I don’t care if I kill the partition table and give a nonsensical label - there maybe purists who aren’t happy with this.

There are probably countless pitfalls that might happen, such as wrong drive selected in multiple drive systems resulting in the wrong drive having its first 512 murdered - that isn’t an issue for me but at least it defaults to N

@Wayne-Workman
/thread hijack /offtopic /sorry.

Don’t forget that if you ARE running windows at all then hyper-v is (for less than 20 guest servers assuming a 2 host cluster) a better option, you get TWO virtual licenses for each server license purchased. thus a “free” host per hyper-v instance. If you were to license 10 virtual MS server guests on 2 hosts using centos as a clustered host then you would need to purchase 20 server licences (and you cant use enterprise licences as you aren’t licensing the host in this case). If you were using a pair of server 2012R2 hosts and had 10 server guests then you only need to purchase 10 licences as each license is worth 2 thus each host has 5 licenses each thus 10 total clustered servers.

And FWIW it is an ABSOLUTE NIGHTMARE to license desktop windows in a virtual environment - long story short don’t do it. Normal licenses do NOT cover W7 being installed in a virtual environment you need specialist licences (VDA). Seems mad? Yes it is.

also if you are creating a “universal” image and can spare an extra 10gb or so then look at “snappy driver” for deploying drivers. That is what I use for my “universal image”.

for our Windows 7 image I did the following:

*set up our “unattend.xml” using a workstation.
*from within a hyper-v guest, install fresh W7 and boot into audit mode.
*install base software (office, serif, any other site licenced products).
*install FOG client
*copy snappy drivers and download the packs.
*configure setupcomplete.cmd to run my snappy driver .bat. This BAT autoruns snappy driver to install MISSING drivers, then deletes the folder of drivers when completed.
*windows update
*copy the unattend.xml created earlier.
(I then “backup” my VM at this point so I can come back to it later)
*sysprep shutdown and generalise
*capture the image into FOG then restore VM from backup ready for the updates next time.

To get snappy drivers to install, simply open it, update the packs, then in your setupcomplete.cmd call

SDI_x64_Rxxx.exe -autoinstall -autoclose -license

The batchfile also cascades into an RD and shutdown reboot at the end of it so it is self cleaning. NOTE I also shutdown the FOGClient service in my setupcomplete.cmd as I DONT want it to join the domain as soon as it sniffs the network! It will join the domain (and rename etc) AFTER the drivers have been installed and rebooted. The batch file also ensures our KMS key is added and I run slmgr /ato. I also add the office KMS key and activate that at the end too, whilst there are default KMS keys when products are installed, I have other images with MAK keys as some of our machines are destined to be unplugged from the network (exam machines etc).

My image is a monster, weighing in at 40gb but works on every single machine I have thrown it on and they are pretty much ready to go (WSUS always has more updates on first boot)

A note that clonezilla works using the method.

well as far as “stable” goes ive been using build 5370 for 3 months. Ive rolled out a couple of hundred windows 8.1 images - all of which are unattended from lan boot -> AD joined “ctrl-alt-del” screen (I use snappydriver via a script on OOBE boot).

I guess later trunk have other fixes but since that build worked for me (and works for me) I have no reason to update yet. I’d give trunk a test in your lab and decide for yourself; it is a world away from 1.2 - if only for the speed increase (not quite double the speed but I’d be surprised if you don’t notice it)

@wayne-workman just a note (apologies for gravedig).

I never got round to upgrading FOG over the summer and TBH I might look at building a brand new FOG server to supplement the old one.

However, we are rolling out windows 10 1709 and am STILL using FOG 5370 with the legacy client. Everything works just fine as before (I cannot remember when I installed 5370 but it is quite a few years ago now…)

yes it can do what you want. FOG stores all of its data in an easy to read database. We looked at automating FOG from another of our applications in the past (but didn’t bother going forward as it didn’t actually save us much admin). For FOG tasks (scheduled, instant, delayed etc), simply insert something into the fog.scheduledTasks database for a scheduled task etc (the fog.taskTypes gives you an indication of what you can do). The backup process kicks in from the pxe by default in FOG. You can set a task to reboot the PC too in FOG if you wish.

Obviously it is a DISK BACKUP not an incremental backup. Expect the backups to take many minutes if you run high compression. Even low compression is a heavy task.

It will be a network backup, local backups (to USB etc) will not be possible unless you configure multiple storage nodes per PC (not practical).

If you can SSH to your NAS then go ahead and try (after backing up of course)! Quite a few NAS do support an awful lot of things; looking at my own synology, that has a dashboard where you can install web servers, MySQL, bulletin boards - all sorts really so it isn’t too much of a stretch to imagine FOG can be installed on the NAS. It all comes down to what flavour of Linux the NAS is running and how friendly it is to repo’s vs recompiling…

thank you

We are using Fog old trunk 5370. It has worked on everything I have thrown at it and it has a few of my own hacks in there (murdering partition tables on fresh never used drives for example). I am looking at moving to 1.4.0 but I have one reservation, is the fog client enforced? We still use the legacy client as I never had much success with the non-legacy version. There are many machines on our network using the legacy client.

@george1421 yes. That way the fog server considers /images local so the “defaultmember” pie chart works just fine.

Dell XPS 11 (2x USB3.0 ports and wifi - no onboard LAN) just imaged successfully with a LAN7500 USB NIC. I tried all sorts of NIC without success, both with the BIOS, UEFI and various ROM-o-MATIC USB configurations. In the end a good old USB2 startech USB21000S2 worked from the BIOS (legacy) without issue. I didn’t try via UEFI.

Inventory was OK. Imaging windows 10 dropped twice but picked up again (“kevent 2 may have been dropped”).

Posting just in case others are trying to handpick USB NIC

@george1421 I “cheesed” the dashboard in our FOG by mounting the image folder on the FOG server as NFS from my synology NAS (rather than add another storage node). Using an 815+ here and have been for a while.

well as far as “stable” goes ive been using build 5370 for 3 months. Ive rolled out a couple of hundred windows 8.1 images - all of which are unattended from lan boot -> AD joined “ctrl-alt-del” screen (I use snappydriver via a script on OOBE boot).

I guess later trunk have other fixes but since that build worked for me (and works for me) I have no reason to update yet. I’d give trunk a test in your lab and decide for yourself; it is a world away from 1.2 - if only for the speed increase (not quite double the speed but I’d be surprised if you don’t notice it)

old topic but I had a similar issue, I simply created a DHCP reservation for the problem laptops (I only had a few) and set the 67 for those laptops on the reservation. That way I could leave fog alone and let the DHCP service give the appropriate details out (and being a reservation it is a permanent solution as far as I care).