Hey folks, title says it. but here's the background. I have a set of valve index knuckles controlers. absolutely -l- -o- -v- -e- them. But valve has screwed us. Cant actually buy them now, and the Right one has died. Not even like, dead dead, but the spicy pillow inside has become spicy.

Yes, ive attempted a repair. If it went well i wouldnt be here. Even for an experienced guy like me, this is not an easy swap, and the damn battery connector on the board snapped off the PCB while i was working on getting the mother flipping GLUE off of the connection. WHY.. WHY VALVE?! This was after I tried to see if i could get valve to replace or repair, and no good. Im out of options.

Ebay has em, but i dont trust that they arent just going to have the same problem.

What are my alternatives?

FOLKS! ITT is about to hit its 200th episode. In the past 200 episodes we've given you all hundreds of hours worth of linux knowledge, career advice, good demos, and maybe a few bad ones (we wont talk about that sudo episode), and even a number of stories about our own histories, and that of Red Hat.

When we hit 100, we made the terrible decision to give you our top 100 linux commands! And almost 2 hours on-stream later, we did it! You can see it here: https://youtube.com/live/8GYtQymteJQ

We'd like to do something fun for our 200th, and we're looking for your ideas! What do you think Scott and I should subject ourselves to for the 200th episode? Whatever it is, it needs to be RHEL related.

So comment below! What do you think we should do for 200? Maybe you'll get a mention on the show if we use your idea!

Thanks!

How to build your career in technology is a moving target, and it's moving quickly! In this special career-focused episode of Into the Terminal, Nate and Scott step away from the command line to have a candid conversation about building a career in Linux in 2026. What certifications matter? Does your distro choice count? And how exactly does AI play into the future of sysadmin roles? Watch the full episode here or build your skills in the interactive labs.

Key Takeaways for 2026

1. Certifications vs. Degrees

Do certifications get you the job? According to Scott (who hires technical talent), certifications don't get you a job—they get you an interview.

• Hiring managers look at certifications to establish a baseline of knowledge. Performance-based certifications (like Red Hat's RHCSA and RHCE) are highly valued because they prove you can actually execute the tasks on a live system, rather than just answering multiple-choice questions.
• While some large organizations (or government entities) might still rigidly require a 4-year degree for management tracks, certifications are often a fraction of the cost and, when paired with experience, can open the majority of doors in the industry.

2. The Power of Home Labbing and Projects

If you can't get on-the-job experience doing the thing you want to do, build it yourself! Nate shared how he originally built his own DNS, mail, and web servers just to learn how they worked, which gave him tangible projects to point to in interviews.

• Pro Tip for Interviews: Don't just talk about your successes! Be prepared to discuss your biggest failures. Hiring managers want to hear about how you messed up, how you recovered, and how you documented the fix so the team never steps on the same rake twice.

3. Does Your Distro Choice Matter?

Short answer: Yes, to an extent. While it might be fun to run a niche Linux distribution that only a handful of people in the world use, it won't help you much in a corporate job interview. Stick to the big players (RHEL family, Ubuntu LTS, etc.). The enterprise world runs on these major distributions, and the skills you build on them will translate directly into your career.

4. The Role of AI in Sysadmin Careers

Is AI replacing sysadmins? Rather than fully replacing roles, expect to see smaller teams heavily augmented by AI tooling. "AI does Linux really well," but organizations need practitioners who can securely prompt the AI, vet the scripts it generates, and evaluate the output. You are the layer between the AI and the real world.

• The Ultimate Study Hack: Use AI as your personal study buddy! Spin up a Linux VM and tell a generative AI agent to log in and break something on the system without telling you what it broke. Then, use that broken system to practice real-world troubleshooting and diagnosis.

Links & Resources

• Video URL: Into the Terminal 198 - Building your Linux career in 2026
• RHEL Developer Registration: https://developers.redhat.com/register
• Interactive Labs: https://redhat.com/interactive-labs
• Discord: https://red.ht/rhel-discord (Note: This now points to the Fedora community discord which hosts the RHEL channels!)
• Nate Lager: https://social.undrground.org/@gangrif
• Scott McBrien: https://www.linkedin.com/in/scott-mcbrien-349b356

Into the Terminal is a show dedicated to helping you grow your knowledge of critical administration skills for Red Hat Enterprise Linux. Whether you are new to Linux or new to RHEL, join us for a hands-on look at commands, processes, and tools.

Ever find yourself wasting time manually spinning up virtual machines for your home lab, only to spend 10 more minutes updating them right after deployment? In this episode of Into the Terminal, we walk through tools and tips to streamline and automate virtual Linux machine deployments on Linux. Watch the full episode here or test your skills in the interactive labs.

The Essentials

Building virtual machines doesn't have to be a manual chore. By combining a few utilities that come with libvirt (or the libguestfs-tools package), you can easily automate your lab infrastructure.

Advanced Features

Building the Perfect Disk Template with virt-customize

Instead of installing every guest OS from scratch, using templates saves a significant amount of time. You can use tools like virt-customize to offline modify your disks before booting them up. In his deployment scripts, Nate uses it to create a standard baseline for his lab:

• --root-password password:secret: To standardize the root password across new deployments.
• Modify GRUB timeouts: E.g., increasing the default 0 timeout for easier troubleshooting.

Pro Tip: While virt-customize can do almost anything (inject SSH keys, timezone info, and run arbitrary commands), you might want to keep the disk level simple and let provisioning tools handle runtime specifics.

Provisioning the VM with virt-install and cloud-init

Once the disk is ready, virt-install instantiates the VM. Nate walks through a few critical flags in his script that make automation seamless:

• --import: This is a vital flag when working with a template. It tells virt-install the disk already has data on it and not to expect an empty drive with a PXE boot or ISO file.
• --autoconsole none: Ensures you aren't immediately trapped in a terminal-based console right as the script runs, allowing you to gracefully check on it later using the Cockpit web console.
• --cloud-init: Nate passes a cloud-init file to automatically inject his personal user, SSH keys, and sudo access during boot. You could do this at the disk level, but grouping user-specifics in cloud-init separates the user data from the base template cleanly!

Kernel Tuning with tuned-adm

Scott breaks down the tuned daemon, which dynamically updates the running kernel (by making live changes to /proc and /sys without requiring a reboot).

• Run tuned-adm active to check your current performance profile, or tuned-adm list to see available options.
• When tuned detects that RHEL is running as a virtual machine (via BIOS info), it automatically assigns the virtual-guest profile.
• This profile is based on the throughput-performance profile. It increases the writeback percentage (so the VM caches more disk writes before syncing back to the hypervisor) and lowers swappiness (making the system less interested in using swap space), keeping your VM running smoothly.

Maintaining Fresh Templates with Image Builder

The problem with templates is they age quickly. Deploying an old template (e.g., RHEL 10.0) means you'll have a massive stack of updates waiting for you upon boot. Instead of updating old VMs manually and running virt-sysprep to make a new template, Nate recommends utilizing the Red Hat Hybrid Cloud Console and Image Builder to create a brand new, fully updated disk image.

With Image Builder, you can define a custom blueprint, inject necessary packages (like tmux and vim-enhanced), define custom partition tables, and download a slim, fully up-to-date .qcow2 image (often just ~1.1GB for a basic install). Simply place the new image in /var/lib/libvirt/images/ and update your scripts to use it!

Quick Reference Card

# 1. Clone a base image
cd /var/lib/libvirt/images/
cp ~gangrif/rhel-10.2.qcow2 ./cloneme-10.qcow2
chown root:root ./cloneme-10.qcow2

# 2. Customize the disk offline
virt-customize -a ./cloneme-10.qcow2 --root-password password:secret

# 3. Provision the VM using the customized disk and cloud-init
virt-install --name my-new-vm --memory 2048 --vcpus 2 \
  --disk ./cloneme-10.qcow2 \
  --import --os-variant rhel9.0 \
  --autoconsole none \
  --cloud-init user-data=/path/to/cloud-init.yaml

# 4. Check Tuned Performance Profile
tuned-adm active

Links & Resources

• Video URL: Into the Terminal 193 - Linux Virtual Machine Tips & Tricks
• RHEL Developer Registration: https://developers.redhat.com/register
• Interactive Labs: https://redhat.com/interactive-labs
• Nate Lager: https://social.undrground.org/@gangrif
• Scott McBrien: https://www.linkedin.com/in/scott-mcbrien-349b356

Into the Terminal is a show dedicated to helping you grow your knowledge of critical administration skills for Red Hat Enterprise Linux. Whether you are new to Linux or new to RHEL, join us for a hands-on look at commands, processes, and tools.

Applying an update with dnf is the easy part. But what happens before and after you run that command? This week on Into the Terminal, Scott dives into the full lifecycle of a Critical Vulnerability and Exposure (CVE) mitigation. From discovery and embargo periods to CVSS scores, assessment, and fleet-wide rollout strategies, we cover what it really takes to manage security vulnerabilities on Enterprise Linux.

Watch the full episode: https://youtube.com/live/WHcgvoszf-M

Try it yourself on a live lab: https://redhat.com/interactive-labs

The Mechanical Part — Applying the Update

At the end of the day, mitigating a CVE usually involves updating a package. For example, if a vulnerability is announced for buildah, the mechanical fix is straightforward:

dnf update buildah

However, depending on the package, you may need to take additional steps:

• Standalone binaries (like buildah): The next time the command is run, the updated version is used.
• Services (like httpd): You must restart the service to load the patched version into memory.
• Core components (like the Kernel or glibc): A full system reboot is required.

Understanding CVEs, CVSS Scores, and RHSAs

When a vulnerability is discovered, it goes through a lifecycle before you ever run dnf update.

The Embargo Period

When a security researcher finds a vulnerability, they typically report it to the vendor first. The time between this private report and the public disclosure is the embargo period. This gives vendors like Red Hat time to assess the issue, develop a patch, and have it ready for customers the moment the vulnerability is made public.

CVSS Scores and Severity

Vulnerabilities are scored using the Common Vulnerability Scoring System (CVSS), which rates them from 0 to 10 based on factors like ease of exploit and potential impact (e.g., gaining root access vs. reading a restricted file).

These scores translate to severity levels:

• Low
• Moderate
• Important
• Critical

Note: Red Hat often provides a Vendor-Adjusted CVSS Score. Because Enterprise Linux is compiled with specific security flags and architectures, a vulnerability that is a 9.0 on a generic Linux system might only be a 7.5 on RHEL due to built-in mitigations.

Red Hat Security Advisories (RHSA)

When a patch is ready, Red Hat issues an RHSA. This document details:

• The CVEs addressed
• The specific products and versions affected
• The exact updated package versions required to mitigate the issue

The Mitigation Process — Fleet-Wide Response

When a major vulnerability hits the news (like Dirty Pipe or Copyfail), your security team will immediately ask: "What is our exposure?"

1. Discovery: Identify which systems are running the vulnerable software. You can do this manually via Ansible/SSH, or automatically using tools like Red Hat Satellite (formerly Insights), which tracks the packages installed across your registered fleet and highlights vulnerable hosts.
2. Assessment: Determine if the vulnerable software is actually exposed in your environment (e.g., is the vulnerable service running and accessible?).
3. Planning: Formulate a rollout plan. For critical infrastructure, this might involve taking nodes out of a load balancer, updating and restarting services, and putting them back into rotation to ensure zero downtime.
4. Application: Run the updates and verify the remediation.

Quick Reference Card

# Update a specific package to mitigate a CVE
dnf update <package_name>

# Example: Updating buildah
dnf update buildah

# Remember to restart affected services!
systemctl restart <service_name>

Links

• Watch the episode: https://youtube.com/live/WHcgvoszf-M
• Get started with RHEL: https://developers.redhat.com/register
• Try it on interactive labs: https://redhat.com/interactive-labs
• Nate Lager: https://social.undrground.org/@gangrif
• Scott McBrien: https://www.linkedin.com/in/scott-mcbrien-349b356

Into the Terminal is a weekly livestream covering critical administration skills for Red Hat Enterprise Linux. Whether you're new to Linux or new to RHEL, join us for hands-on looks at commands and processes, ask questions, and grow your knowledge.

I'd love to slowly replace all of the Google home/nest devices in my house with HA Voice.

One of those is the nest hub with the screen. It does info displays but mostly it's a nice integrated photo frame.

Has anyone built something similar using HA Voice as the back end for voice commands?

Two weeks ago on Into the Terminal, we went deep on vim — from the essential commands every sysadmin needs, to multipliers, macros, visual blocks, and customizing your .vimrc. Whether you're brand new to vim or you've been using it for years and just never learned beyond :wq, this one's for you.

Watch the full episode: https://youtube.com/live/UfzjXbLVwwo

Try it yourself on a live lab: https://redhat.com/interactive-labs

The Essential ~15 Commands

You can be minimally proficient in vim with about 13–15 commands. Everyone's list is a little different — here's what our two hosts use, and where they differ.

Getting In and Out of Insert Mode

Moving Around

Pro tip from the show: You can also open a file at a specific line number from the shell:

vim +20 myfile.conf

Handy when your compiler tells you there's a syntax error on line 20.

Deleting / Cutting

Copy (Yank) and Paste

Note: dd puts the deleted line into a buffer, so it doubles as "cut." This terminology predates the modern "copy/paste" — back when vi was created, the terms were yank (copy) and delete (cut).

Undo

In vim improved, u walks back through your full edit history. In classic vi, it only undoes the last command. Also note: undo history is per-session — if you quit and reopen the file, the history is gone.

Save and Quit

Note: :w! won't help on a read-only filesystem. That's enforced at the filesystem level, not the permission level.

Advanced Features

Multipliers — Do Anything N Times

Prefix any command with a number to repeat it.

Line Anchors — ^ and $ in Combination

The ^ (beginning) and $ (end) characters aren't just for movement. You can combine them with other commands:

By learning just ^ and $ as anchors, you get multiple delete behaviors without memorizing separate commands for each.

Search

Search and Replace (Substitution)

Reading Files and Command Output Into Your Buffer

The :r !command trick is a time-saver — no need to run a command, redirect to a file, then :r that file.

Security note: :! lets you run any shell command. If someone has sudo access to run vim as root, they can shell out from vim with full root privileges. Use the NOEXEC option in sudoers to prevent forking from within escalated sessions.

Saving to a Different File

Great for making a backup before committing changes, or for cleaning up data and writing the result to a new file while preserving the original.

Visual Block Mode — Comment Out Code in Bulk

1. Press Ctrl+V to enter visual block mode
2. Use arrow keys (or j/k) to select the lines you want
3. Press I (capital I) to insert at the beginning of the block
4. Type your comment character (e.g., #)
5. Press Esc — the character appears on all selected lines

This is how you comment out a block of code in seconds. Works great for Python, shell scripts, config files — anything with a line-based comment character.

Pasting Pre-formatted Code Without Mangling

Vim's auto-indent is great when you're writing code, but it wrecks pre-formatted code that you paste in from your clipboard.

When paste mode is on, you'll see -- INSERT (paste) -- at the bottom. This is much easier than the alternative of typing :set noautoindent nosmartindent nocindent.

Macros (Named Buffers) — Record and Replay Commands

1. q + a letter (e.g., qt) — start recording to buffer t
2. Perform your commands (move, delete, edit, etc.)
3. q — stop recording
4. @t — execute the macro stored in t
5. @@ — re-execute the last macro
6. 100@t — execute macro t 100 times (multipliers work here too)

Use :reg to view all your named buffers and see exactly what commands are stored in each one.

Example from the show: We had a CSV file with 114 lines that needed the same edit on every line. Instead of doing it by hand, we recorded a macro that moved down a line, repositioned the cursor, deleted the unwanted data, and then replayed it across the entire file with 100@t. Seconds instead of minutes.

Customizing Your Session (and Your .vimrc)

All of these can be run as one-off commands with :set, or persisted by adding them to your ~/.vimrc file.

RHCE exam tip from the show: When you sit down to take the exam, put :set cuc and :set number in your .vimrc first. The cursor column highlighting will save you when editing Ansible playbooks and YAML files.

The Pattern for Set Commands

• Turn something on: :set feature
• Turn something off: :set nofeature
• Some settings use = syntax: :set tabstop=4
• Some commands like syntax don't use set at all: :syntax on

Quick Reference Card

GETTING AROUND           EDITING                    SAVE/QUIT
i    insert mode         dd   delete line           :w   save
I    insert at line start D    delete to end of line :q   quit
a    append after cursor  d^   delete to line start  :wq  save & quit
Esc  back to normal      yy   copy line             :q!  quit no save
/    search forward      p    paste                 :w!  force save
?    search backward     u    undo
G    end of file         x    delete character
:20  go to line 20

POWER MOVES              SETTINGS (in :set or .vimrc)
5dd    delete 5 lines    tabstop=N     tab width
Ctrl+V visual block      expandtab     spaces not tabs
qt...q record macro t    hlsearch      highlight matches
u/t     play macro t      cursorcolumn  column highlight
100@t  play macro 100x   number        line numbers
:%s/a/b/g  replace all   paste         paste mode
:r !cmd    read cmd output syntax on   syntax highlighting

Links

• Watch the episode: https://youtube.com/live/UfzjXbLVwwo
• Get started with RHEL: https://developers.redhat.com/register
• Try it on interactive labs: https://redhat.com/interactive-labs

Into the Terminal is a weekly livestream covering critical administration skills for Red Hat Enterprise Linux. Whether you're new to Linux or new to RHEL, join us for hands-on looks at commands and processes, ask questions, and grow your knowledge.

Hello,

I have a seerr deployment, with, of course, radarr/sonarr behind it. Works wonderfully, movies or series that i ask it to track, get found and show up like magic.

But, i have one series im trying to grab, that it just doesnt seem to like.

"The Vampire Lestat", If i search for it from the overview, it finds the series, but then when i try to request it, seer comes up and says its unable to automatically match the series, even though it showed it in my search results. And it gives me 6 cards for other series with similar names. None of them are the correct series. But if i go download the show manually, and throw it into my media server, Seerr happily shows me the episodes that I have, and the ones that are missing.

I'm just not sure where to go with this, sure i can manually download them, but i'd much rather let seerr/sonarr do their job.

Any suggestions?

I switched! Last night I finally got my cellular service moved over to my new Pixel 9 pro running grapheneOS. From Iphone.

One thing I wasnt able to thoroughly test until I switched, is android auto. I followed the instructions to get the modified android auto app installed from the Graphene app store. Set it up to use a usb connection rather than wireless. Today I got to try it, and it works!

I have a few questions though.

First, voice. I did not, and do not want to, install google's assistant. I think that should feel obvious. However, one of the benefits of Android Auto or Apple Carplay is its ability to let me hands-free interact with the media on my device. Or even just read a text message. Without a voice assistant installed, it seems that this is not possible.

Second is Maps. I also didnt install google maps. I found two alternatives, that seem like theyll work well, and they work with android auto. But when i enable turn by turn directions, they speak, but... its like they're reading out an API call rather than the actual directions. I wonder if this is because the app is supposed to be sending something like a call to the google voice assistant, but instead its doing its best to speak them because there isnt one.

So my qustions are two. First, is there an alternative voice assistant folks like? Could the google assistant be used without it being a privacy nightmare? And second, is my read of the weird text im getting from turn by turn directions, does it stand to reason that adding a voice assistant would fix that?

Thanks!

Man. It feels like im coming home. I was a long time android user, switched to iPhone a few years ago because I was sick of googles free spying.

I was never really happy with apple. But i felt like android just wasn't an option. But it seems graphene has matured enough that its absolutely a viable daily now!

So, the actual question. I have an apple watch. Mainly for the health monitoring. I use it for workout tracking, sleep tracking, and importantly heart monitoring. I won't go into details but I'll say that the ability to record a ECG on the fly is important to me.

The pixel watch does this from what I'm reading. So does Garmin. Are there others?

Is running a pixel watch just giving google a window back into my life? Are they as cavalier with health data as they are with other data?

Thanks for any input.

I am in the process of moving from Plex, to Jellyfin. Its going pretty dang good. I have one thing thats not perfect yet.

Audiobooks.

In plex I used a plex audiobook library, with the Prologue app on iPhone, and it worked really damn well. On Jellyfin I found that it supports audiobook libraries, and on iPhone i found the Plappa app, which works really damn well (practically identical to Prologue).

But I have to throw a wrench into things. I am also on a bit of a digital self relience/rights journey, and decided to switch back to Android, on Graphene (because I am sick of feeling like the device in my pocket isnt actually built for me, but for companies to gather data on and profit from me). So I would love an app that works like Plappa for Android.

I know i know, everyones going to recommend audiobookshelf. I'm not opposed to it, but i feel like im so close here, and would really rather not manage yet another app.

Thanks!

Hello there folks,

I am experimenting with Local LLM's, on a Framework Desktop, Ryzen AI 128gb model.

Ive got litellm, and ollama running, with a few different models.

This week I decided to see what I could do locally with OpenClaw, and its not going well.

I keep running into issues with speed, and the chat's sort of losing the thread, i think because of context sizes.

Some troubleshooting with Claude (chatting with claude, not using openclaw with claude) makes me think that this just isnt possible, because Openclaw sends a massive prompt to the llm for each task, and its just taking my local deployment forever to deal with it, and its losing things like, what tools it has at hand.

Anyone else doing this?

I'm using qwen2.5:32b-instruct for my model, let me know if you need more info.

Thanks.

So, for reasons that have to do with ideology i've left Audible behind, and now buy all of my audio books through Libro.fm. All of Jim's books are there! Win!

But out law isn't and as im reading more about why, it sounds like its an audible exclusive.

Im curious if anyone knows if thats a permanent thing? Or if I'll eventually be able to buy it elsewhere? It sounds like The Law was originally an Audible exclusvie too, and that's available on Libro.

Im considering just buying the dead tree edition, so I can read it, and then ill pick up the audio book later (i really love Marsters performances), If its ever available on anything except Audible.

Thanks!

Bash has a lot of shortcuts, a lot of them aren’t obvious, but you pick them up over the years. Watch Scott and Nate share their favorite command line tricks and shortcuts on this episode of Into the Terminal.