Few questions regarding CentOS in medium size studios


Hi guys,
Currently have CentOS 7 boxes where we manually install the nvidia driver with dkms. While this works, not sure how scale-able this will be (currently a small studio, about to expand rapidly), should we need to update drivers on 30+ boxes, i would really not want to do this manually. How are other studios managing the drivers? manually with the run file, or via repo (elrepo/negativo17)?.

Also would be helpful to hear from anyone who manages multiple Centos boxes, how you manage them? Do you use spacewalk + ldap + kerberos? Also any insights into cifs vs nfs?

Any help would be greatly appreciated


are the systems similar in hardware?.. you should aim for this and use image deployment methods.

perhaps look at Kickstart here: https://www.centos.org/docs/5/html/Installation_Guide-en-US/ch-kickstart2.html


Use a repository like ELRepo for Nvidia drivers. Don’t use the Nvidia installers unless you’re using them to make your own packages. Most studios use configuration management software to manage Linux machines (Chef, Puppet, Ansible, and my personal favourite Salt). Look at realmd for identity management if you haven’t already. CIFS versus NFS depends on many factors so more information would be needed.