Not Hijacking this post, but just wanted to “upvote” this and say I am having the same exact problem with the same exact model , Lenovo T14 Gen 2. I narrowed it down to what you guys have found so far before finding this post (that the nic driver in the kernel is most likely the issue). Using a lan adapter as a workaround, but if I find anything I more I will let you guys know.
Dustin
@Sebastian-Roth Just want to say this is resolved. Not sure why, but installing on Debian fixed the issue. We tried the same exact thing with Ubuntu and CentOS on a reinstall importing the same exact database but couldn’t resolve it…I know as a tech person that doesn’t make sense as to why Debian would fix the issue…but it did and has been working for a few weeks now solid.
Dustin
@Sebastian-Roth Sorry I did not see your response before my “update” post, but thank you for the reply. I tried changing that on the other installs that we had (the 14.04, 16.04 and CentOS), but it the unresponsiveness stayed. The weird thing is that we have not changed the client checkin time on the Debian install and it is still working great. Does Debian handle Fog better in general?
Just a quick update…We installed on Debian and imported the database and it is working… Web GUI is responding much better as of now, but we have not changed anything except the OS it is installed on. We have no idea why it is all the sudden working but it is. So a reinstall on Ubuntu 16.04 had the same issues, CentOS has the same issues, but Debian works. I will update later today if it is still working well.
Dustin
Version - 1.5.4
VM - 8GB Ram
A few days ago our Fog server stopped acting normally. The web GUI was super slow, almost unresponsive. We tried everything we knew to figure out why, and even re-installed to see if that would fix the issue. Our original server is Ubuntu 14.04 where the problems started happening. We tried to re-install fresh on 16.04, same problem. We also tried installing on CentOS and got the same issues. The web GUI might work for a short time, but eventually it will fail or just “spin” and do nothing. If the web GUI does work, trying to start an image task or going to the storage node section are things that cause it to just “spin”. Logins take forever…etc… I have tried searching the forums and doing things like https://forums.fogproject.org/topic/10717/can-php-fpm-make-fog-web-gui-fast to no avail. If I run “top” typically we see lots of php-fpm processes. I also tried updating to trunk to see if that fixed the issue, but same results (super slow, unresponsive GUI). Maybe we have royally screwed something up along the way and did not realize it, but we have not been able to solve the issue. If someone has ideas, just let me know, any help is appreciated.
Dustin
This isn’t really related to FOG specifically, but I have seen that others have posted Windows 10 issues that have to do with imaging in here (so I thought I would try). I have made many Windows 10 images, and have never ran into the issue that I am currently experiencing with any of the past builds (1511, 1607). Basically the issue I am having is after deploying my image (Windows 10 - Version 1703), the administrator auto-logins indefinitely when it should only auto-login once. The auto-login option is in the unattend.xml file that is ran with sysprep. I have my auto-login count set to “1” currently, but it will re-login to the administrator profile multiple times which is causing issues. I haven’t changed anything in the way I do things, and it has always worked before (which leads me to believe this is a 1703 issue)… Let me know if you have any suggestions or if you too have been experiencing this. I haven’t been able to find a fix or even anyone experiencing the same issue. Thanks!
Dustin Olson
@Sebastian-Roth @Wayne-Workman Just an update, making a folder for the image, switching the type, renaming image file to sys.img.000 and copying partition info into the new folder made the image work.
@Sebastian-Roth @Wayne-Workman Sorry for the late reply, I was out of the office until now. They were already set as partimage images. Good catch Sebastian, yes they were Vista images that were actually Windows 7 images (old images that were pre 1.2 when one had to use Vista to upload Win7 images). I changed the image type to Windows 7 from Vista, and since the image file was not a “folder”, I created a folder for the image, moved the image file into the folder and named it “sys.img.000”, and copied partition files from another Win7 image and put them in the folder with it. The image is downloading right now, but I am not positive it will work with me hacking at it haha. Is there a better way to recover those old images or should my process work? (Ill update this if it does or doesn’t work). Thank for the replies, I appreciate the help and feedback!
Dustin
I update my images on VM’s, and recently I have ran into an issue where I get the error message “Could Not Remove Old Parition” when trying to download a few images(pic below). I saw some posts on the forum with the issue, but I didn’t see an answer for something I should try. The images I am trying to download are Win7 images uploaded with partimage not partclone (I believe, as they are not folders on the storage node). Any ideas on how I can troubleshoot this? Let me know what you are thinking when you have time, thanks!
Dustin
@Tom-Elliott I see. Ya it should have at least told me what interface after the “from”, didn’t realize that until now. Thanks Tom!
hmm… That’s strange. It’s eth0 and I can ping stuff. I just told it “Yes” ,and typed in eth0 to continue. Didn’t mean to do an unnecessary post, just wanted to make sure something wasn’t wrong.
I have been updating my storage nodes to the latest version, but just ran into a odd issue where during the install script, it will not accept “No” as an option for changing from the default network interface. It accepts yes, but not the “No” Option… Below is the picture of my shell. Has something changed in the install script? (Also tried going directly to the server and running the install script, same results). Should I re-download the repository and try again?
Hi All,
I hope I am not asking a question that is obvious or already been addressed elsewhere, but I just upgraded my main Fog install (the server itself) from 1.2 to 1.4.0-RC-9.3, and I am wondering if there is any known issues with upgrading the nodes along with the server? I haven’t had any problems pushing images or anything, but before I upgrade the nodes I thought I should check to make sure there is isn’t anything I need to be aware of that I might have missed before I decide to run the update on all the nodes. Thanks!
Dustin Olson
@george1421 Thanks so much for your response. Just to share with anyone else who might use it, I just created a Powershell script that gets the hostname and emails me an alert about the system. Script below if you want it…(I am calling the script if certain criteria are not met (like missing drivers or something)…
$test=Test-Connection $env:COMPUTERNAME -count 1 | select Ipv4Address
$IP=$test.IPV4Address.IPAddressToString
$MAC=(gwmi -Class Win32_NetworkAdapterConfiguration | where { $_.IpAddress -eq $IP }).MACAddress
$Hostname=Invoke-WebRequest http://192.168.64.14/fog/service/hostname.php?mac=$MAC
$Hostname=$Hostname -split("\n")
$Hostname=$Hostname[0]
$Hostname=$Hostname.Replace("#!ok=", "")
$Hostname=$Hostname.Trim()
Send-MailMessage -To "YOURTOEMAIL" -From "FOG_WARNING@FOGALERTS.NET" -Subject "$Hostname Alert" -SmtpServer "YOUR SMTP SERVER" -Body "$Hostname had a problem... You need to fix this!!!!!"
Mod edited to use codebox.
This might seem like an odd question, but I want to grab the hostname of the computer from the FOG database before the computers hostname actually get’s changed on the Windows client side (at this point the FOG service hasn’t changed the Windows hostname yet). So in other words, once the computer has been imaged and the initial scripts have started, from the Windows client side, I want to log the hostname somehow if something specific happens. For example, I have a way to detect if the video driver has been successfully updated. If the driver has not been updated during my driver install process, it would be nice to take that hostname and LOG it or email myself using scripts. I started to look a little bit into it, and the solution was leading me down a path of connecting to the mysql database and running a command using putty against the hosts table to get the hostname. But that seemed extremely messy, so before I go any further, does anyone know of a way to do this using the existing FOG service or something from the Windows client side? Any custom command I can run from the Windows FOG client side to grab data like that? I hope that makes sense.
Dustin Olson
@Tom-Elliott I will eventually (and want to). The reason for reservations on upgrading Fog right now is because I am not wanting to fight through all the things that can break when upgrading my live system (not saying it would, just there’s a chance). I work for a school district so I would rather test things before-hand to ensure the system is still functional if it is needed.
@Tom-Elliott Ok. I will just leave this as is then unless there is something you guys want me to do. Thanks for all
@Tom-Elliott No, not on my live fog system. This has fog 1.2. Is the RC S40 script still applicable if I am able to get it?
@Tom-Elliott I installed Fog1.2 on another system and copied over the init.xz to my current fog setup. After sending the image, it just zipped through to a black screen again 
. So it did the same thing that it did before. shoot…
@Sebastian-Roth Below is the S40network contents from my original init.xz…
#!/bin/sh
case “$1” in
start)
echo “Starting network…”
/sbin/ifup -a
;;
stop)
echo -n “Stopping network…”
/sbin/ifdown -a
;;
restart|reload)
“$0” stop
“$0” start
;;
*)
echo “Usage: $0 {start|stop|restart}”
exit 1
esac
exit $?
@Sebastian-Roth Yes, I named it “4.6.2” in “/var/www/{,html}/fog/service/ipxe/”. I am positive the kernel file is named correctly and that it loads when PXE booting. Below are the screenshots you requested.
