Installing a new clean fog server
-
- VM Host = leave network settings on dynamic
- VM Guest = configure network settings on static
- FOG Client = configure network settings on static
So if we set it up like this all 3 would have different ip addresses correct?
-
@Predator said in Installing a new clean fog server:
FOG Client = configure network settings on static
This IP address would need to be assigned by dhcp since you need pxe boot information (provided by dhcp server). If you want fog client to be locked at a static IP address then use dhcp reservations. But either way fog client must be assigned by dhcp.
And you are correct all three need different IP addresses.
-
Thanks, i hope i will not encounter too much problems during installation fingers crossed.
-
@Predator You should not as long as you follow the networking rules.
If you have issues, post back here and we can help you. Getting clients to pxe boot is sometimes trouble. Its not a fog issue but a network environment/infrastructure one.
-
@predator said in Installing a new clean fog server:
What is the best and simplest OS to use with FOG? I have read that some have problems with ubuntu… Furthermore i am a noob in Linux distro’s and it would help me a lot to have something that’s stable and simple with a GUI.
My goal is to implement a virtual fog server for the enterprise where i am currently working so i can deploy images over pxe boot.
I’ve been using Ubuntu Server 16.04.2 without any issues, and it’s super easy to deploy and manage. Also, I know this might seem scary or counter-intuitive, but I would actually recommend NOT using a GUI (that’s why I deployed the “server” version of Ubuntu).
I’m a little confused by the fact that you said your goal is to “implement a virtual fog server” in an “enterprise” environment, but you’re trying to build this “environment” using VMware Workstation. Have you considered deploying VMware ESXi? It’s free and would be more ideal than using VMware Workstation. I’m not saying you can’t do it that way, but it’s not a very common way of going about it, especially in an “enterprise” environment.
With that said, if you’re going to use VMware Workstation, I would think you would want to setup the PC that’s running VMware Workstation (the IP of this machine can be DHCP or static), then build an Ubuntu VM using NAT and set a static IP on this guest.
If you have a DHCP and DNS server in your environment (which it sounds like you already do), during the FOG setup, you should be able to make the following selections when installing:
-
When prompted if you’d like to change the default network interface, type “N”
-
When prompted if you’d like to setup a router address for the DHCP server, type “Y”
-
When prompted for the IP address to be used for the router on the DHCP server, press Enter if it already shows your correct router address (or type in the correct one if it’s not already displayed)
-
When prompted if you’d like DHCP to handle DNS, type “Y”
-
When prompted for what DNS address DHCP should allow, press Enter if it already shows the correct address (or type in the correct address if it’s not already displayed)
-
When prompted if you’d like to use the FOG server for DHCP service, type “N”
You’ll then want to go to your DHCP server and make sure that the following scope options are configured:
- 066 Boot Server Host Name = make sure this is the IP of your FOG server
- 067 Bootfile Name = undionly.kpxe
You can either configure those scope options globally under IPv4 under “Server Options”, or you can configure them under “Scope Options” under each Scope that you want to allow access to the FOG server.
-
-
Thanks for the information.
The reason why i’m using VMWare workstation is to test it before i put it in our VMWare pool.
Once it’s fully tested and working by me i will put it on a VMWare server and manage it from the vsphere client.
-
@predator To each their own I guess. We do guest testing all the time in our vSphere production environment. You’re the first person I’ve ever heard of who has an ESXi host, but does all their testing in VMware Workstation O_o.
-
Yep don’t mind me it’s just a infrastructural decision.
-
Having problems with entering the following command lines in centOS
echo “Open UDP port 49152 through 65532, the possible used ports for fog multicast”
firewall-cmd --permanent --add-port=49152-65532/udp
echo “Allow IGMP traffic for multicast”
firewall-cmd --permanent --direct --add-rule ipv4 filter INPUT 0 -p igmp -j ACCEPT
systemctl restart firewalld.service
echo “Done.”FYI i’m not using putty but entering the commands in terminal of centOS.
Any idea why this is not working? I get the error command not found…
This is a step of the continue pre-config step from wiki guide: https://wiki.fogproject.org/wiki/index.php?title=CentOS_7
-
Nevermind i managed to make the command lines work.
Can someone explain me which one and why i should activate the following?
Set SELinux to permissive on boot:
sed -i.bak ‘s/^.\SELINUX=enforcing\b.$/SELINUX=permissive/’ /etc/selinux/configSet SELinux to permissive on the fly (this is not persistent, the above config must be done to be persistent):
setenforce 0 -
@george1421 @lmioperations @Sebastian-Roth
Can someone chime in on this?
-
@predator Hi, SELinux must be set to permissive for now in order for certain FOG central services to work correctly (such as NFS).
There have been some attempts to create SELinux profiles instead, but given how often things can change, this is obviously a more cumbersome way of going about it.
You should use both commands. First one will make it so when you reboot, SELinux is in permissive mode, second one sets it in current session so you don’t have to reboot.
-
Simply the
setenforce 0disables selinux (think MS Windows UAC) off without having to reboot your server. Updating the selinux conf file will make the change persistent across reboots. We are recommending you set the value to permissive. This is done so that selinux will record the events but not block them. If you require selinux enabled because of your company’s security policy you can take these logged events and create a selinux profile. This fog selinux profile then can be enabled to allow FOG to run while selinux is enabled. -
@predator said in Installing a new clean fog server:
SELinux to permissive on boot
See if this info helps: https://wiki.gentoo.org/wiki/SELinux/Tutorials/Permissive_versus_enforcing
Looks like George already beat me to a reply :D, and his info ended up being more to the point of what you need with regards to FOG ;).
-
Ok looks like i’m kinda stuck here. I managed to install the fog client and have to visit the management console in my web browser 10.x.x.x/fog/management/index.php
I’m in front of the login page and can’t seem to login. Where the hell did i setup this login i don’t even know…
-
I have setup centos on IPV4 10.2.3.207
When installing the fogclient i choose IPV4 10.2.3.208
I get a response in the terminal to open the management console on http://10.2.3.208/fog/management/index.php
Yet http://10.2.3.208/fog/management/index.php is not reachable…
But when i enter http://10.2.3.207/fog/management/index.php i do get the fog login page…
-
@predator Why did you setup on 207, but then go to 208?
-
@george1421 said in Installing a new clean fog server:
And you are correct all three need different IP addresses.
-
Because George confirmed all three need different ip addresses…
@george1421
1.VM Host = leave network settings on dynamic
2.VM Guest = configure network settings on static
3.FOG Client = configure network settings on staticSo if we set it up like this all 3 would have different ip addresses correct?
-
@predator You have this incorrect.
1.VM Host = leave network settings on dynamic
2.VM Guest = configure network settings on static
3.FOG Client = configure network settings on staticThis should be,
- VM Host = Must be unique, but don’t really care. Not used by/for fog
- VM Guest = configure network settings on static
- FOG Client = dhcp
