Fedora 33 - Workstation Installation: Difference between revisions
DrEdWilliams (talk | contribs) mNo edit summary |
DrEdWilliams (talk | contribs) mNo edit summary |
||
| Line 2: | Line 2: | ||
It is assumed that either the Fedora Workstation (Gnome desktop) or KDE Plasma spin is used as the base distribution and the desktop is configured and functioning properly. | It is assumed that either the Fedora Workstation (Gnome desktop) or KDE Plasma spin is used as the base distribution and the desktop is configured and functioning properly. | ||
== System Configuration == | == System Configuration == | ||
Revision as of 22:05, 29 November 2020
After doing the Basic Fedora 31+ Installation, these are the steps required to get the workstation up to general functionality.
It is assumed that either the Fedora Workstation (Gnome desktop) or KDE Plasma spin is used as the base distribution and the desktop is configured and functioning properly.
System Configuration
In addition to the basic installation:
- Mount the /workspace and/or /shared Ceph filesystems (see Ceph Storage Cluster)
- Mount the /files and /backup filesystems from storage1 (NFSv4 mount)
- Configure (or verify) the network devices:
- WilliamsNet (10.0.0.0/24)
- StorageNet (10.1.0.0/24) - optional if needed (if device is available or VLAN device definition)
- iSCSINet (10.2.0.0/24) - optional if needed (may require VLAN device definition
- CUDA + NVIDIA drivers
General Utilities
A collection of tools that are useful in a workstation; some may already be installed, but just covering the bases:
- Cockpit web-based system administration
- Google Chrome
- use the repo-based installation on the Google instructions page so you get updates
- make it the default browser by searching for 'Application' in the KDE settings app and selecting it for the default web browser.
- pdsh - parallel shell
- see pdsh article in ADMIN Magazine
- in fedora repo:
sudo dnf install -y pdsh pdsh-rcmd-ssh
- by default in fedora, only the rsh module is installed ??
- Synergy (see Synergy Installation
- Java
- either the openjdk in the fedora repo or the 'real stuff' from Oracle
- Oracle Java 8 is required to support java web start applications (like the EquaLogic Group Manager tool)
- Gnome Terminal -- it's just better than the KDE konsole app, though konsole has gotten a lot better
sudo dnf install -y gnome-terminal
Communications and Collaboration Tools
There are a lot ... but these cover much of the landscape:
- Microsoft Teams
- download client from https://teams.microsoft.com
- default install will establish a teams repo to facilitate updates
- Slack
- download client from https://slack.com/download
- Zoom
- download client from https://zoom.com
- Kopete (Jabber client)
- in fedora repo
- VMware Horizon Client (for access to SAIC VDS)
- download from SAIC VDS webpage
Development Tools
- VS Code
- download client at https://code.visualstudio.com/download
- will create a repo to get updates
- if possible, copy ".vscode*" directories from an existing configuration
- LOTS of plugins/modules/addons/extensions ... easy to waste a lot of time configuring
- PodMan -- rootless docker replacement ... in the repo
- Docker-CE -- is this even available?
Extended/Optional Stuff
This is not intended to be a complete list of all the random/useful stuff I do on workstations, but just a catch-all for links and reminders
Enable RPM fusion
If not already done as part of the NVIDIA/CUDA driver debacle, install it -- there are useful things there
sudo dnf install https://download1.rpmfusion.org/free/fedora/rpmfusion-free-release-$(rpm -E %fedora).noarch.rpm sudo dnf install https://download1.rpmfusion.org/nonfree/fedora/rpmfusion-nonfree-release-$(rpm -E %fedora).noarch.rpm
Elastic Stack
Still trying to decide what to do about this ... but it may replace the log aggregation server in the production kubernetes cluster. Installation instructions on their website (https://elastic.co), if you can dig through all the stuff. Specific things I've tried so far include:
- metricbeat - general metrics collection
- filegbeat - file scraping and general text-based data collection
- logstash - collection and processing pipelines to aggregate and send to elasticsearch
- elasticsearch - data storage, indexing, and analysis platform
They offer a repo to pull packages from -- it is kept current with GA versions.
InnoDB
Time Series Database -- in some way equivalent to ElasticSearch, but not as versatile with the whole platform. Very good at metrics collection, but not so much at text collection.