A true diskless environment with remote home dir for users

[hsafe]

Hi Steve;
Back in 2012 I did a working env of the DRBL, now my requirement is not imaging over LAN or a terminal from the server. But the ability to give the thins a graphical environment with their remote home dirs which include personal chrooted dirs saved at the server.
An example could be like a puppy linux + some customized apps inside this image plus the home dirs of users which are saved at the server side. The application of this project could be for offices that maintain a high level of cost-control and efficiency while users data is not lost and the source image is loaded from the server to the diskless clients.
Reading recently found out that there is little regarding the users dir and how to mount them in the working env of a projected image to the thin client. Can you provide me some guideline and direction
Very much appreciate it. Thanks

[stevenshiau]

@hsafe,
What did you mean by "how to mount them in the working env of a projected image to the thin client"? Could you please describe more? In DRBL environment, we just use NFS for that.

Steven

[hsafe]

Hi Steve
Thanks for the reply. You see I have already set up the server and client environment. Now all the settings are working properly and as expected. Let me explain more:
-The environment is a KVM virtualization with 2 node server+client. Server is configured properly and assign dhcp and push the clonezilla spash to the client. So far so good...
-While configuring the drbl I accepted the default image location which was /home/partimage
-I have also downloaded the image: drbl-live-xfce-2.3.1-6-i686-pae.iso(not exactly know the purpose of it honestly)
-My setup environment is the full DRBL mode environment+ users have their own home dirs on the server(no disk at their end)
Now the silly question is what image and on what order should I do it? Which dir should I use to place the bootable presumably live disk in?Where to define the users accounts ?

[hsafe]

Hi again,
now even I can't get pass the tftpboot stage after some reboots. Initially I could.The firewall started which I took care of (stopped it)and the service conf in xinetd for tftpd is ok as well. My host name was named like localhost-ethernet4 which now is changed to drbl.
The tftp process status:
It starts with tftp://myserverIP/pxelinux.0 ...........connection time out (http://ipxe.org/4c126037)
I did not give any dns server for the set up and accepted the default.
If I tftp to the localhost and try get I can see the list of the files in /tftpboot dir

[hsafe]

Hi Again
I managed to fix the above issue as under RHEL-based linuxes multiple interfaces cause issue on which GW to be picked up as default and in my case wrong interface. Now all the DGW is gone I will pass the tftp load but in the end am faced with the issue:
tftp prefix:
unable to locate configuration file

Here is a bug report of the server config fyi:
https://oss.borhancid.com/owncloud/public.php?service=files&t=0ae0afc6f58b93ead0f36db40f1a2561

[stevenshiau]

Did you check the log file on the server? Maybe you can find the clue why clients can not find the configuration file.
BTW, did you disable selinux on your CentOS server?

Steven