@Tom-Elliott

Very good, fantastic!
Re-insalled using the existing installfog.sh - did its thing (checked packages, database scheme etc.)

cd /root/fogproject/bin ./installfog.sh

WOL and postdriver scripts seem to be working again!
Thanks! I shall run this command below so it doesn’t happen again until I am prepared to get Ubuntu updates.

apt-get -y remove unattended-upgrades

Which, after running the install script again, worked fine. Thanks, @Tom-Elliott

@Tom-Elliott When I said I changed from zstd to gzip, I changed it in the image settings via the GUI, pictured below. Then I recaptured & redeployed and it worked fine.
0_1493855732085_zstd and gzip.png

@najeebrehman df -h will show you all of the mounted devices. lsblk will also show you the mounted (connected) devices, just in a different layout. Look for a device that has the same size as your usb flash drive. Then follow tom’s commands below substituting the drive (/dev/sda, /dev/sdb, etc) that is identified as your usb device.

RESOLVED: I was forcing https on the forward and as soon as I removed that it started working and I have already had 3 machines check in and they were the ones I just installed the client on remotely.

Thanks Tom. Just the switch on the AD Username did the trick. Thanks for your help so quickly…

James

@george1421 Windows was telling me after the second boot (inaccessible boot device) this was after the bootloader.

Regards X23

@Seydoo A “hash was the same no matter the snapin” leads me to think it was failing to get the file. When this would fail, it would cause the file to be nulled, so the hash being returned was that of a null value. (for all cases, which is why it was the same).

@garysaxon I sure would like to see the results of the following commands.

ls -la /images
ls -la /images/<image_name>

And then a screen shot of the image definition in the fog management gui.

Ok, so it looks like the default image engine was changed. Once we got the images added to the system and changed them from PartImage to Partclone Gzip they started deploying correctly.

Thank you Tom!

@Tom-Elliott
That did it. After that last update, I was able to deploy the raw image successfully! Thanks for working with me on this.

@fogusers That really doesn’t answer the question what happens after stopping web services failed? Does it return you to a command prompt?

Also in the …/bin directory (where you ran the ./installfog.sh command from) there is a log directory and in that log directory there is an install log with error and the version you are installing. Post that error log here for review.

@Jaymes-Driver

Late response, I know, but you’re right. I think the Ubuntu updated itself and reverted a change I had made with the link below. I had to go back in and blank out the SQL password again:

Is this Ubuntu 16.04? See here: https://wiki.fogproject.org/wiki/index.php?title=Ubuntu_16.04#MySQL_password_behavior

@george1421
thanks for pointing me in the right direction. I reran the installer with wifi internet and static ethernet and dhcp is working now. I think during the initial install i was trying to use two connection profiles on ethernet, rather than utlizing wifi card for internet to complete install and ethernet for fog only.

The following fixed it:

https://forums.fogproject.org/topic/7135/boot-pxe-issue-check-if-mntcheck-exist/2

@Rama This thread is fairly old, and the OP is probably not going to respond. Can you please create a new thread with all the details, fog version, how you are doing it, etc.

@Amerhb Is the fog server doing dhcp or getting dhcp? Did you change the fog server’s IP address? We need a lot more information, what you’ve stated doesn’t have enough information for us to even know what the problem category is.

@grant1842 said in Your database connection appears to be invalid.:

I had to re run the installer as well as remake my dnsmasq 2.76. … but my dnsmasq reverted back to 2.75 which was causing the boot file to not be received correctly.

This is only speculation, but I’m guessing that the linux distro you use updated dnsmasq from (2.72 or 2.73) to 2.75, thus overwriting your compiled 2.76. As wayne said the package manager only knows about what it does. Hopefully the next time it happens is when your distribution updates to 2.76 or newer. Then it won’t matter at all.

On the plus side you were able to discover and fix the problem pretty quickly. So good job there.

@george1421 It’s ubuntu, so ps aux | grep apache2 but yeah he probably got slammed by Canonical’s nasty automatic MySQL update.

@fredlwal that’s basically what a reinstall does.