Difference between revisions of "Cocalc on Fedora — installing and debugging"

From Wiki4Intranet
Jump to: navigation, search
m (StasFomin moved page Cocalc on Fedora to Cocalc on Fedora — installing and debugging without leaving a redirect)
Line 13: Line 13:
 
** Actually, I spent several weeks trying to fix regular crashed of «cocalc-docker» on podman, and failed.
 
** Actually, I spent several weeks trying to fix regular crashed of «cocalc-docker» on podman, and failed.
 
* You want to share your postgresql server (with centralized backups) with other services.
 
* You want to share your postgresql server (with centralized backups) with other services.
* You want to fast experimenting cycle with updating packages, patching configs/code inplace without rebuilding/reinstalling large docker blob.
+
* You want to fast experimenting cycle with updating packages, patching configs/code inplace without rebuilding/reinstalling large docker blob, or remapping container filesystem.
 
* You like old-school configuration management with Vagrant/Ansible.
 
* You like old-school configuration management with Vagrant/Ansible.

Revision as of 05:52, 12 December 2021

How to install Cocalc on Fedora.

Why?

NIH?
  • Yes, there exists https://github.com/sagemathinc/cocalc-docker based on Ubuntu.
  • May be you did not like Ubuntu and prefer Fedora, because of
    • Reliable SAT-based dnf
      • Or what to use some packages existing specifically on RPM world.
    • Your target server or your desktop already on Fedora (or both) and you did not what to use 30GB docker blobs, and what to share large cocalc dependencies (TeXLive, etc) between several services (MediaWiki, Overleaf, … ).
  • You are tired with docker/podman, swarm/kubernetes/… hell, and want to use classic systemd orchestration.
    • Actually, I spent several weeks trying to fix regular crashed of «cocalc-docker» on podman, and failed.
  • You want to share your postgresql server (with centralized backups) with other services.
  • You want to fast experimenting cycle with updating packages, patching configs/code inplace without rebuilding/reinstalling large docker blob, or remapping container filesystem.
  • You like old-school configuration management with Vagrant/Ansible.