Post Download Scripts - Not Executing
-
For this image, what do you have defined for “Image Type” and “Partition”? Perhaps you have something misconfigured here which is why the UUID of the SWAP space is not set properly?
I’d guess that you should have “Multiple Partition Image - Single Disk (Not Resizable) - (2)” and “Everything - (1)” respectively.
-
Oh. My. Gosh. I have been using Single Disk - Resizable this whole time! I thought this was the correct configuration, one DRIVE, which is resizable for each partition. When did this change, out of curiosity? When I first downloaded the FOG Project, I swear the multiple partition selection did not exist. This has me wicked excited! I was just about to go home too. Time for one more test!
-Dustin
-
@mrayzies You came to the same conclusion I did (as I was finishing mowing the grass tonight, and still thinking about this issue). The fstab being empty is on the FOS Engine linux OS, that I can understand. Your mount command is needed to connect to the target’s local hard drive (which is not mounted automatically by FOS). We have to do this if we want to tweak the unattend.xml in the windows realm of the post install scripts.
-
Testing this today - had to leave work last night. I am hoping this will solve it. It feels promising, heh.
-Dustin
-
@dholtz-docbox said in Post Download Scripts - Not Executing:
Server
- FOG Version: 1.3.0-rc11
- OS: Ubuntu 14.04
Description
I am near the end of what I need from the FOG project, and everything seems to be functioning pretty solid so far. However, I am having issues kicking off post download scripts.
I have followed the following two articles to no success…
- https://forums.fogproject.org/topic/7740/the-magical-mystical-fog-post-download-script-under-construction/3
- https://wiki.fogproject.org/wiki/index.php/Auto_driver_Install
My /images/postdownloadscripts/fog.postdownload looks like…
#!/bin/sh . ${postdownpath}update_swap_uuid.sh
And update_swap_uuid.sh looks like…
#!/bin/bash ORIGINAL=$(grep -v '#' /etc/fstab | grep swap | cut -d ' ' -f1 | cut -d '=' -f2) CURRENT=$(blkid | grep 'swap' | cut -d ' ' -f2 | cut -d '"' -f2) TIME=$(date +%Y%m%d%H%M%S) if [ "$ORIGINAL" = "$CURRENT" ]; then exit fi # NOTE: Backup 'fstab' cp /etc/fstab "/etc/fstab.$TIME.backup" # NOTE: Overwrite ORIGINAL with CURRENT swap UUID sed -i -e "s/$ORIGINAL/$CURRENT/g" /etc/fstab
I have had no issues using the update script itself, so I am not sure why FOG would have an issue with it.
I have also given ‘755’ rights to the whole postdownloadscripts/ directory.
Am I forgetting something or doing something incorrectly? I have tried little permutations of editing the fog.postdownload and update scripts, as well as hunting for settings and permissions to allow access to these contents.
Also, where would I look for any indication that these are being run? I have looked very closely through the entire process and have seen no output text about running this particular region of scripts.
Many thanks for your time!
-Dustin
You can change your update_swap_uuid.sh to something like:
#!/bin/bash # First we need to get the partitions of this disk (typically $hd in resize or single disk nonresize) getPartitions "$hd" # Now to iterate the parts for part in $parts; do # Print a nice message dots "Mounting partition $part" # Attempt the mount mount $part /mnt >/dev/null 2>&1 # It didn't mount, inform and start at top if [[ ! $? -eq 0 ]]; then echo "Failed to mount the partition" continue fi # Test to see if fstab exists on this part # If not un-mount the directory and print the message for the user and start from top of loop if [[ ! -f /mnt/etc/fstab ]]; then umount /mnt echo "Done, fstab not on this partition" continue fi # File was found echo "Done" # Let the user know what is happening dots "Checking and updating swap UUID" # Get the fstab uuid currently setup ORIGINAL=$(grep -v '#' /mnt/etc/fstab | grep swap | cut -d ' ' -f1 | cut -d '=' -f2) # Get the current real partition swap CURRENT=$(blkid | grep 'swap' | cut -d ' ' -f2 | cut -d '"' -f2) # Get the current date/time TIME=$(date +%Y%m%d%H%M%S) # Check if the original is the same as the current if so # if so, unmount and inform start at top if [[ $ORIGINAL == $CURRENT ]]; then umount /mnt echo "Done, UUID unchanged" continue fi # NOTE: Backup 'fstab' cp /mnt/etc/fstab "/mnt/etc/fstab.$TIME.backup" # NOTE: Overwrite ORIGINAL with CURRENT swap UUID sed -i -e "s/$ORIGINAL/$CURRENT/g" /mnt/etc/fstab >/dev/null 2>&1 # If the sed fails inform, un-mount, start at top if [[ ! $? -eq 0 ]]; then echo "Failed to update file" umount /mnt continue fi # All succeeded inform and unmount echo "Done, UUID updated" umount /mnt done
-
Setting the image to “Multiple Partition Image - Single Disk (Not Resizable) - (2)” did not work either.
I guess what stumps me is why I am having this issue, or what I need to do to avoid it. There isn’t even anything special about the drive I am imaging, it’s just a fresh installation of Ubuntu 14.04 Desktop. How do you guys generally configure your swap partition? Linux needs this swap partition, so it is awkward to me that I am not seeing more resources on this issue while having received it so easily.
I am trying to avoid the post download scripts solution at the moment, and am looking into other ways of configuring the system before imaging. At the moment, I have tried just using the physical name in lieu of the UUID for the swap drive and am testing this as I write this.
I will continue posting my progress.
-Dustin
-
Oh, that’s awesome! I am still learning how to do more with the system through shell scripts, so this is very cool to read through. I will update my current script to reflect this and give it a whirl. It might not be the best solution, but I can figure out another solution if this one works for this particular milestone.
-Dustin
-
@dholtz-docbox I’m only looping the found partitions to make things a bit more dynamic. But if you know the FOS system recognizes the disk as /dev/sda and you also know the root etc/fstab will be on partition 1, you can forgo the loop and simply mount the /dev/sda1 and make your edits directly.
-
@Tom-Elliott : What is the best way for me to learn what I can do in the FOS? I couldn’t locate a primary resource on the FOG wiki regarding it and its available functionality. I wouldn’t mind knowing more about how people use this side of FOG, it is starting to feel like it’s the strongest tool in the whole arsenal.
Also, I was able to capture + deploy the image fine if I used the physical partitions name in lieu of the UUID.
/dev/sda5, none, swap, sw, 0, 0
My primary concern in this solution is the what-if scenario. Should I be concerned about whether this link will change on its own? That would be my primary concern, correct? Someone changing the sym-link between /dev/sda5 and its underlying UUID?
Thanks for going back-and-forth with me on this. I don’t have a lot of people to talk with about topics like this.
-Dustin
-
@dholtz-docbox You’re absolutely right that postdownload scripts is a very powerful tool. It can also be very destructive, but “with great power comes great responsibility…” or something like that.
Essentially, postdownload scripts can be ANYTHING you want them to be. They have the power to iterate over the freshly imaged system to change whatever you may need, however you may need to do it.
Yes, links can change, but chances are you’ll know of those changes the first time you go to work with them. With post download scripts, you can modify EVERYTHING, even HOW the scripts run. For example, on the fog.postdownload you can add if statements to do different things based on whatever you deem necessary. You don’t, necessarily, even need to just source your own scripts, you could just script it directly in the fog.postdownload script. I prefer separating my scripts for modularity reasons.
I don’t mind giving input here and there as it helps EVERYBODY.
-
I had to table these efforts, as I have been unable to successfully clone and deploy an extended linux image without having the swap drive hiccup during the cloning/imaging process. I will return to this as soon as time permits, but I have a few tentative workarounds for this milestone.
-Dustin