@Bastian Looking into this, I can’t believe its still an issue after all of these years (> 5). The only solution I can find is to switch the disk controller to ahci mode from RST. If you have a live system that was installed with RST on windows will hate you. I’ve found instructions on how to fix windows.

The ideal solution would be for you to recreate you golden image and sysprep it on a system with the disk in ahci mode enabled. That way everything will work moving forward.

@Sebastian-Roth So it turns out most of my problems were in fact caused simply by old software. I installed Ubuntu 20.04 and FOG 1.5.9 and it solved all the database updating issues.

Now I just have to solve some DNS problems that I’m pretty sure are not FOG related.

Thanks mate.

@Sebastian-Roth I understand than I was racing my brain on this lol

Turns out the DNS server was not ser correctly and the server got it’s IP via DHCP. We got that fixed.

@jmvela2x Sorry, lost track of this.

Here you have a template for an advanced menu:

#!ipxe :ADVMENU menu colour --rgb 0x00567a 0 || cpair --foreground 1 1 || cpair --foreground 0 3 || cpair --foreground 4 4 || item item1 Do this item item2 Do that item item3 Go back to main menu choose --default item1 --timeout 6000 target && goto ${target} :item1 echo "Do this" || goto ADVMENU :item2 echo "Do that" || goto ADVMENU :item3 chain -ar http://${fog-ip}/fog/service/ipxe/boot.php || goto ADVMENU

@george1421 thank you for all the help yet again 🙂

seems to be all working as my router for the dhcp server and the fog server as the pxe server.

PS. i have to say it is a really nice project/community to flow in like this with people willing to help thnx appreciated.

ok thank i actually install newer version of ubuntu not 20.04 but 19.10 cause i already have this iso,

i install again and i change the dhcp conf with filename “ipxe.efi” and i give you new later.

did as you instructed and seems to be working now thanks 🙂

@Jim-Holcomb I still don’t fully understand. The database is just a point to store and pull the data from. Nothing more.

If you manually rename the folder in Linux, would you expect the database to automatically update when now it has no reference point? It’s the same reason in reverse. Why allow the folder to be renamed in the terminal if it won’t update the database that may be referencing it?

The folder doesn’t get automatically created when you define a new image. This only happens after the image is defined and captured. The pieces handling it are all after the database information is pulled. The GUI is not creating the folder. FTP is.

If you manually change the folder name in the terminal, you will still need to change the folder in the GUI. It’s the same for updating the path in the GUI. While you could temporarily store the change, but what about you are changing the path when the image hasn’t been captured yet?

Disregard - some how the SSH service was offline.

@JDnoble18 download it to the fog server. Run it. That’s it.

How you run is up to you. But I’d say do sh ./renumberImageIDs.sh once it’s on the fog server.

@Pyrol said in PXE-E18: Server response timeout:

I reinstalled FOG and PFSense and it’s still the same.

You are using pfSense as DHCP server. Please share the settings you have in there (picture) as well as the information requested by George.

@Sebastian-Roth Just replaced a mobo on a failed 3500, and it came with latest bios, which now acts the same as the 3510 😞

So it seems my work flow with imaging 3500 only worked due a bug in the firmware on the 3500s.

Going to either set the bios to pass-through disabled on unboxing or put together a new imaging work flow.

@george1421 Thank you. I will delete them shortly to avoid any further issues.

@Sebastian-Roth This is not a new branch, just an upgrade and move from Ubuntu to Centos 7 as it is more stable when upgrading. The only thing I can do as far as getting closer is to move this computer into the server room, but since it is a desktop that would be a little cumbersome. These are juniper switches.

For creating the schools’ base image, I use this same machine to do the work for all of them since it is pretty close to the “golden image” for each one. I am using this Dell Optiplex 7040. I create a legacy image (have some more impoverished districts with old machines) and a UEFI image.

I have not tried different iPXE binaries and I wasn’t aware that you guys wanted me to do a mirror port. I will try and work on this today if I get some time.

Thanks for the guidance.

@george1421 Works great, thanks !

@Sebastian-Roth I chatted with the OP last night and got things worked out. There was a number of things wrong with the setup. We got them worked out one by one.

The IP address for the fog server was incorrect in dhcp In the end FOG wasn’t installed completely on the FOG server. Something happened during the install because the .fogsettings file was missing (not very common so it took us a bit to narrow down).

now with that said linux has a command hwclock that can set the hardware clock to local time. That command does exist in FOS Linux, but it appears to have a dependency that is not met. The other missing command is ntpdate to set the FOS Linux clock to some ntp server on your network.

So in theory we can get to where you need with a little black magic.

ref: https://docs.fedoraproject.org/en-US/Fedora/26/html/System_Administrators_Guide/sect4-synchronizing-date-time-hwclock.html

Never mind, found the option in settings. Now working fine and just have to change the boot device before letting the client do it’s thing.

@gwhiteia Well I can either walk you through updating what you need or you can wait until this feature is available in the dev branch. It should take about 15 minutes of hands on time to setup the patch. You will just have to remember to remove it (one line of code) when 1.5.10 is released later this year.

The decision is yours.