Good morning folks, mad post writer here. I hope you’ll enjoy.
https://docs.nethserver.org/projects/ns8/en/latest/release_notes.html
current version suggest a way for re-install cockpit on the system if needed by the sysadmin. Which works for 3 out of 4 distros supported, while debian AFAIK still sticking to apt rather than dnf
https://docs.nethserver.org/projects/ns8/en/latest/system_requirements.html
System requirements do not explicitly requires nor deny homogeneus underlying OS for deployment. I’m probably wrong here, but AFAIK the cluster (which is currently a still undefined project in the adopter mind) cannot have mixed underlying OS.
Davide Principi in november confirmed that documentation need improvement.
I cannot agree more, NS8 is not beta anymore, avoid issues in design phase should help adoption.
In system requirements, fresh install might be added as mandatory; also should be explicit that the first installed node will irreversebly define all cluster nodes distro (current and future)
Also will define the underlying distro for any node that will be created for restore a cluster node or a container; as far as I can understand, it’s currently unsupported restore a node backup created in RockyLinux 9 on a fresh installed Debian 12 node.
Distribution costrain must be explicit (with some examples if your’re willing to care rule down) in requirements and in every step where something underlying NS8 is involved.
In system requirements no information about partitioning/mountpoints is available. I know that single disk and biggest / partition will be bread and butter for a lot of adopters, however an experienced (or opinionated) linux sysadmin might have differents habits and desires.
Therefore…
https://docs.nethserver.org/projects/ns8/en/latest/install.html
might be a nice idea specify directories used for the containers that will need the most of storage space or the expected partitioning/mountpoint style for the documentation. Or not bury it into Disk usage — NS8 documentation which is 4th… from the bottom.
The first four pages of best practice section could help avoid a lot of initial setup and design issues from less experienced adopters. I’d link them at the end of the system requirements page empasizing that might be really useful to read for a smoother first ride.
The disk usage link could be also provided into this page
https://docs.nethserver.org/projects/ns8/en/latest/backup.html
with a note like “paths and files location of NethServer 8 are available here”
Experienced sysadmins would find it useful for creating scripts or second tier backup processes “because Murphy was a clueless optimistic engineer”.