First Impressions / Introduction
-
Just recently got my foot into the door of corporate IT work a few months ago. Currently working as Jr. Systems Admin and I was tasked with creating a new Symantec Ghost server. Long story short, Ghost kicked my ass. Couldn’t figure it out and got to a point where I had no idea where I went wrong and decided to start over. I used FOG to image computers at a prior internship and decided to take a shot at creating my own server.
I’ve been playing around with it all weekend in my virtual test environment using Virtual Box. Windows Server 2012 R2 hosting a basic Domain Controller, pfSense to create a virtual WAN connection (so I don’t mess with my roommates connections), a few end clients, and now FOG 1.2.0 on an Ubuntu 14.04 install. I’ve successfully registered hosts, uploaded an image, deployed said image. I absolutely cannot wait to start testing this in my production environment. Two large warehouses, roughly 400-500 computers in total. My senior team members have never heard of it and want a basic walk through / presentation this week.
After some initial work, I would love to have every end client in both buildings registered to their respective FOG server and be able to image a computer without leaving my desk. I do have admin privileges for both buildings. What are some issues I should keep in mind while implementing this into our system? I’ve read that ensuring switches are enabled to support multicast is key, we have 6 IDF’s in each warehouse running Extreme Network Switches. I’m pretty new to writing script files, FOG seems like an excellent place to practice. What type of simple Snapin’s would be valuable for basic automation and learning? Any help would be greatly appreciated!
The active posts from developers and community support will certainly have me floating around these forums a lot more.
-
For valuable snapins, I would probably start with basic cmd scripts, and then play around with silently installing some exes / msis. (For an exe installer the silent parameter is usually /s or /S).
I may be a tad bias when it comes to this, but I would recommend trying out our development builds on another vm / server. You’ll find imaging speeds are significantly faster, and there are some added security benefits. You should also check out the new beta client (my baby) which works with our development builds. It’s far more responsive than the 1.2.0 client, has quite a few security patches, and is more … ‘professional’ in areas, such as shutdown prompts and such.
(See https://wiki.fogproject.org/wiki/index.php/Upgrade_to_trunk) -
Welcome to both the Fog community and the Corp IT community! Also going to throw in my usual suggestions to go with Hyper-V if you are already using a Windows Server environment, and to go with Debian instead of Ubuntu for server installs.
-
Hey mtmuch and welcome from a fellow FOG aficionado and recent sysadmin. Been using FOG off and on since the 0.29 release and particularly in the lastest 1.2.0 and upwards dev releases, FOG rocks… The team behind it are awesome too and very responsive.
Another thing to watch out for with newer FOG versions is depending on how old your environment is to ensure portfast is enabled on your switches - we’re still waiting on one of our larger sites to have this turned on.
I’ve had success with snapins created using .msi files and several with .exe where the installer silent switches are well documented. Others have been a bit hit and miss but the snapin system itself works very well.
You’'ll need to visit each desktop I’d imagine to set them to boot from nic for PXE booting unless they are set that way already. If they are and you have some sort of inventory system (we use Lansweeper) you might be able to upload all the hosts and mac addresses via a csv file rather than having to manually register them.
I’m also a fan of the new plugin system which has been integrated. We have remote sites and the location plugin solves an issue of replicating between main server and remote storage nodes. Also liking the pushbullet plugin for alerts sent to your phone.
There is a bit of work involved in getting it setup in the beginning, but once you have it configured and clients talking, it will save you so much time. Zero-touch is the buzzword you want to mention to your senior team members.
Enjoy!
-
Appreciate the quick replies guys! I was able to play around with FOG in my production environment today. Ubuntu Server 14.04 & FOG 1.2.0, smooth sailing registering a few clients and uploading an image from a Dell Optiplex 755 running Win 7 Pro. Ran into some issues deploying said image onto another Optiplex 755 with . Host would PXE boot to FOG menu, recognized that I had assigned it a download task for the image. Once partclone screen appeared, I was received image is not a partclone image and the computer would reboot to PXE, rinse and repeat.
I attempted some suggestions that you guys provided like upgrading to trunk using the wget methold, I also updated my kernel to the recommend version suggested for the Optiplex 755 on the wiki. Now the system won’t boot into FOG menu when the host is set to run a task. It’s very likely I did something incorrectly as I’m pretty new to linux.
I’m considering doing a rebuild of the server on Dediban tomorrow like you gentlemen have suggested. Practice makes perfect right?
-
@mtmulch Sorry I haven’t responded sooner. Busy day at work for me. While I try to keep up with bug reports and what not, I am by far from perfect and hopefully any issues you encounter please post them. I may not be perfect, but when I can replicate a problem, or have a directly line to a problem, I am generally very quick to try to get them patched. I try to maintain a relatively working version even in the development versions of FOG. That said, there may be periods where things are FAR from stable.
I would like to welcome you to the community and hopefully the prospect of learning, and questions we may ask if you’re encountering problem is not too much to scare you from using FOG into the future.
-
@mtmulch I advice against using wget to keep up-to-date with dev builds. We release several new dev versions daily, and wget will not work well for keeping up to date. I recommend using git instead.
Just run
git clone https://github.com/FOGProject/fogproject.gitAnd a fogproject folder will be made with the latest and greatest. Then just run
git pullinside of the fogproject folder whenever you want to update the installer.
-
Noted Jbob.
I’ve been doing some research for the past hour or so. With trunk updates is it still necessary to sysprep a Windows 7 / Vista host before you upload an image? We run both. Ideally I’d like to create a master image for each type of workstation we have. Probably 5 different models. Based on what I have read would sysprep be the way to go? I’m going to discuss our license policy with the team. Lots of the machines have been Frankenstein-ed from what I can tell. Same model, different amounts of RAM and varying HDD sizes. Where does FOGprep fall into place with sysprep?
Any best practice suggestions would be extremely helpful.
-
@mtmulch
Welcome to FOG! Don’t hesitate to ask questions, we are here to help. And hopefully you’ll find some way of giving back to the project. See my signature for details.
@need2 said:
go with Debian instead of Ubuntu for server installs.
Or Fedora 22.
https://wiki.fogproject.org/wiki/index.php/Fedora_21_Server@Kiweegie said:
You’'ll need to visit each desktop I’d imagine to set them to boot from nic for PXE booting unless they are set that way already. If they are and you have some sort of inventory system (we use Lansweeper) you might be able to upload all the hosts and mac addresses via a csv file rather than having to manually register them.
Newer systems will allow you to set firmware settings over the network, I hear. I haven’t done it myself yet but I know it would save me a ton of foot-work in the future.
@Kiweegie said:
Zero-touch is the buzzword you want to mention to your senior team members.
Yup.
For some reason, people at my organization want to walk around hitting F12 to network boot. I could care less about making my job difficult for “Job Security”. Part of my goal in I.T. is to make my job easy - and to make my replacement’s job easy, and any less than that isn’t fair or right or moral to not only your employer but yourself. You should always push yourself to find better ways to more efficiently manage more systems at once - how did the Enterprise Administrator who is responsible for 20,000 computers get into his position? It darn sure wasn’t by walking around hitting F12 every time imaging needed to happen - or by walking around manually uninstalling one Antivirus just to install a different one on 500 computers.So yes, the buzzword is Zero-touch, “I’ll make EVERYTHING Zero-Touch, and the next guy can walk in behind me and easily pick up the ball due to my simple naming conventions, ample & well written documentation and resource citation, and well-configured infrastructure” … is what you should really be aspiring to.
Please help us build the FOG community with everyone involved. It's not just about coding - way more we need people to test things, update documentation and most importantly work on uniting the community of people enjoying and working on FOG!
Daily Clean Installation Results:
https://fogtesting.fogproject.us/
FOG Reporting:
https://fog-external-reporting-results.fogproject.us/ -
@Wayne-Workman said:
Or Fedora 22.
https://wiki.fogproject.org/wiki/index.php/Fedora_21_ServerNice Job
-
Welcome to the world of FOG!
Your question about sysprep/fogprep and different models seams to be to complex to be answered in general. Here are some hints and ideas:
- Different amounts of RAM shouldn’t be any trouble.
- For varying HDD sizes you can use “Single Disk - Resizable” as image type (should work with Win7/Vista, AFAIK).
- sysprep/fogprep see here: https://forums.fogproject.org/topic/2403/fogprep-vs-sysprep-windows-7-64/4#
We run FOG (Win7/10 clients) without sysprep or fogprep perfectly. It always depends on what you want to have. Read up on what sysprep is actually used for to see if it is of help in your case!
-
I run sysprep on images that go on multiple types of machines, and don’t on model specific images.
oh, and i also run my server on Ubuntu
(though i’ll probably change distro’s on my next major rebuild) -
Gentlemen, I solved the issue I was discussing earlier but ran into another problem. Corporate does not want a FOG server on their production network as of now, so I was instructed to take it down.
With that being said, I started building a lab. Got a spare computer as my FOG server and a 24 port switch. Now my question to you guys is how can I install the trunk version of FOG without an internet connection during the install process? Is this possible?
I found this article, https://wiki.fogproject.org/wiki/index.php/FOG_on_an_Isolated_Network.
If I am reading it correctly, I allow my one and only NIC on FOG server to accept it’s IP via DHCP for the initial install. Then during the “What is the IP address to be used by this FOG Server?” phase I assign it a static IP for my lab environment. Obviously I don’t want to hook it up to our production environment and accidentally put a second DHCP server on the network so I’m a bit skeptical at this point in time.Any advice would be greatly appreciated!
-
What I did was setup on a server with two NICs and used it as a gateway with it handling fog and DHCP on the NIC for the lab and pulling the internet in on the other so fog stays isolated from the production network. If you just want the server hooked up during the install look here https://wiki.fogproject.org/wiki/index.php/FOG_on_an_Isolated_Network and if you want to try my config you might start with this document I put together from a website I can’t seem to find at the moment. Private FOG Server Config.docx
-
@mtmulch I would recommend just putting it on the network, setting it up, and then just remove it from the network. This way is much more simple.
Just tell FOG you don’t want to use DHCP.
Then, later on, just manually setup the DHCP. it’s easy, we have lots of examples here.
-
So I’ve got the lab setup, more issues I’ve ran into. I can successfully PXE boot into the FOG menu but, once I get there if I select anything the machine does a complete reboot.
I tried changing my dhcp settings to load unidonly.kkpxe no, same results. So I am currently unable to register a host. I have no problem accessing the web GUI but I cannot register the host there either.I am currently running the latest trunk release as of 9 am this morning.
-
@mtmulch Please let us know the exact model of computer you’re trying to network boot. Also try another model too. You might also take a look at our working hardware list: https://wiki.fogproject.org/wiki/index.php/WorkingDevices
Let us know, we are here to help.
-
Sorry for the vague responses. The first model I had issues with was an Dell Optiplex 780, second model I tested was HP Compaq 4000. Looking at that list the HP model is not supported, while the Optiplex is. I see a recommended kernel for the Dell, could this potentially be the culprit? Not sure if this helps at all but I was able to successfully get these to register through the FOG boot menu and web GUI using FOG 1.2.0 on Ubuntu 14.04 ( my current trunk install is running on this).
Is it possible that this is a result of my DHCP configuration? I’m currently using isc-dhcp-server is dnsmanq a preferred option?
-
@mtmulch said:
Sorry for the vague responses. The first model I had issues with was an Dell Optiplex 780, second model I tested was HP Compaq 4000. Looking at that list the HP model is not supported, while the Optiplex is. I see a recommended kernel for the Dell, could this potentially be the culprit? Not sure if this helps at all but I was able to successfully get these to register through the FOG boot menu and web GUI using FOG 1.2.0 on Ubuntu 14.04 ( my current trunk install is running on this).
Is it possible that this is a result of my DHCP configuration? I’m currently using isc-dhcp-server is dnsmanq a preferred option?
There is no preferred way to deliver the boot file option, what matters is that the option gets delivered. Some methods are more feasible and possible than others for different people and different situations.
Just because a model isn’t listed in the working hardware list doesn’t mean it isn’t supported. That list is only what people have graciously taken the time to tell us about - FOG is a 100% community driven project. We rely on people like you to tell us things.
Set your boot file to undionly.kkpxe in your isc-dhcp conf file, then restart DHCP, and then see if you have better luck.
-
@mtmulch said:
I can successfully PXE boot into the FOG menu but, once I get there if I select anything the machine does a complete reboot.
Really any entry you select will cause the machines to reboot? memtest, boot from disk, …? All?
I have no problem accessing the web GUI but I cannot register the host there either.
Any error here? Why wouldn’t you be able to add hosts by hand in the web interface?
I am currently running the latest trunk release as of 9 am this morning.
I just saw Tom talking about a buildroot issue he fixed today. Maybe you get the newest latest trunk again and see if that helps?
