Welcome back to The Deep Dive, your go-to source for cutting through the noise,
Download transcript (.srt)
0:00
0:03
0:04
0:09
0:09
0:14
0:17
0:18
0:22
0:22
0:24
0:33
0:36
0:38
0:42
0:43
0:44
0:47
0:51
0:52
0:56
0:57
0:58
0:59
1:00
1:01
1:03
1:06
1:11
1:11
1:16
1:20
1:22
1:26
1:26
1:29
1:32
1:36
1:37
1:38
1:39
1:44
1:45
1:46
1:48
1:51
1:52
1:53
1:56
1:59
2:00
2:01
2:02
2:05
2:09
2:13
2:15
2:16
2:17
2:18
2:19
2:20
2:24
2:24
2:27
2:31
2:32
2:33
2:34
2:37
2:39
2:42
2:43
2:46
2:48
2:51
2:52
2:54
2:58
2:58
3:04
3:09
3:12
3:13
3:15
3:20
3:20
3:23
3:27
3:28
3:32
3:34
3:35
3:39
3:41
3:45
3:45
3:46
3:51
3:52
3:55
3:56
3:58
3:59
4:00
4:03
4:09
4:10
4:11
4:12
4:18
4:18
4:21
4:23
4:25
4:28
4:30
4:31
4:32
4:33
4:36
4:38
4:39
4:41
4:43
4:45
4:48
4:54
4:56
4:57
5:01
5:07
5:08
5:09
5:13
5:14
5:15
5:16
5:19
5:23
5:24
5:25
5:26
5:27
5:28
5:31
5:32
5:35
5:37
5:40
5:41
5:43
5:44
5:45
5:48
5:49
5:53
5:53
5:54
5:55
5:56
5:57
5:58
6:01
6:05
6:06
6:10
6:11
6:12
6:16
6:16
6:17
6:20
6:22
6:27
6:27
6:28
6:29
6:31
6:34
6:37
6:38
6:39
6:43
6:44
6:48
6:50
6:51
6:54
6:55
6:57
7:01
7:02
7:06
7:08
7:09
7:12
7:14
7:15
7:16
7:17
7:18
7:20
7:21
7:23
7:24
7:27
7:30
7:32
7:36
7:37
7:38
7:40
7:42
7:43
7:44
7:48
7:48
7:50
7:55
7:56
7:57
7:58
7:59
8:03
8:06
8:10
8:10
8:11
8:14
8:15
8:16
8:19
8:20
8:21
8:24
8:28
8:29
8:33
8:33
8:38
8:41
8:42
8:44
8:47
8:48
8:49
8:53
8:53
8:56
8:57
8:58
9:02
9:07
9:08
9:09
9:10
9:11
9:16
9:20
9:21
9:22
9:23
9:27
9:27
9:29
9:33
9:33
9:36
9:38
9:39
9:40
9:46
9:49
9:53
9:57
9:57
9:58
9:59
10:01
10:06
10:10
10:11
10:12
10:13
10:17
10:19
10:21
10:25
10:30
10:34
10:38
10:39
10:42
10:43
10:45
10:46
10:48
10:51
10:52
10:56
10:57
11:01
11:03
11:04
11:05
11:10
11:10
11:13
11:17
11:19
11:21
11:23
11:25
11:26
11:30
11:30
11:34
11:34
11:35
11:35
getting
straight to what matters, and hopefully give you some interesting facts along the
way.
Before we jump in today, a huge thank you to our partner, safeserver.de.
They make this show possible, and they're all about supporting your digital
transformation.
They handle hosting for software, just like the tool we're discussing today, and
generally
help simplify things.
You can learn more at www.deat, that's www.saserver.de, check them out.
Okay, so today we're peeling back the layers on a tool called Bezel.
Heard a bit about this one, seems quite popular.
It's basically a really simple, lightweight server monitoring solution, but it
still packs
a punch.
Docker stats, historical data alerts, all the good stuff.
Exactly, and that simplicity is key, especially if you're maybe just starting out
with managing
your own server, or digital stuff, the whole idea of server monitoring.
It can sound really technical, right?
Maybe a bit intimidating.
Definitely.
Like you need a PhD or something.
Yeah.
But what's great about tools like Bezel is they kind of demystify it.
They're built to give you those aha moments.
So instead of feeling overwhelmed by data, you get clear insights you can actually
use.
It makes the complicated stuff feel, well, accessible, avoids that info overload.
I like that aha moments instead of overwhelmed.
So that's our mission for you today.
Unpack Bezel, look at its main features, and just explain how it makes keeping tabs
on
your digital world less complex.
We want to show you how it makes monitoring approachable.
Whether you're just curious, building something cool, or even managing stuff for a
small team.
Okay.
Let's start right at the beginning then.
For anyone listening, maybe thinking, server monitoring.
What's that?
What exactly is server monitoring?
And why would you even need it?
And when we say server, we don't mean some giant blinking thing in a movie.
Right.
Not usually.
For most of us, it's just a computer that's running something important for you.
Could be your website, an app you built, maybe a personal project you're tinkering
with.
Your digital engine basically.
Exactly.
So monitoring is like the dashboard for that engine.
Think of Bezel less as just a tool and more like your early warning system.
You know, normally if something goes wrong, your site gets slow and app crashes.
You find out after the fact.
You're reacting.
Loading up fires.
Right.
You can use that.
It lets you be proactive.
It gives you visibility into how things are performing, how resources are being
used before
a small hiccup becomes a major headache.
It's the difference between, say, adjusting something minor and spending a whole
night
trying to fix a crash.
Makes total sense.
Avoiding those late night panic sessions is always good.
And that's where Bezel comes in.
Positioning itself as this really accessible solution designed to meet those
monitoring
needs without adding loads of complexity.
You mentioned lightweight and simple.
What does lightweight actually mean in practice?
Does it mean less capable?
Not at all, actually.
Lightweight here means it uses far fewer of your server's resources compared to
some
of the big enterprise monitoring systems, which is a massive plus if you're running
it on something modest, like a small virtual server or even a Raspberry Pi.
Ah, so it doesn't hog the resources it's supposed to be watching.
Precisely.
And simple, well, that's not just talk.
It means the setup is genuinely easy, minimal configuration needed, and importantly,
you
don't have to expose it to the public internet if you don't want to.
It's pretty much ready to go out of the box so you get useful info without hours of
setup pain.
That sounds like a huge barrier removed, especially for learners or hobbyists,
democratizing monitoring
almost.
Okay, let's dig into some key features then, explained simply.
First up, Docker and Podman stats.
Now, for folks new to this, what are Docker and Podman, quickly, and why look at
their
stats specifically?
Sure, so think of Docker or Podman as, like, really neat ways to package your
applications.
Imagine each app lives in its own little self-contained box with everything it
needs.
Okay, like digital Tupperware.
Huh?
Yeah, sort of.
These containers make sure your app runs the same everywhere.
Now, Bezel tracks things like CPU use, memory use, network activity history for
each individual
user.
Ah, the history part is important.
It's crucial because it lets you see if one specific app, maybe your blog or that
script
you wrote, is suddenly going wild with resources.
You're not just guessing that the whole server is slow.
You can pinpoint the cause.
Like knowing which room in the house is using all the electricity, not just the
total bill.
Great analogy.
Exactly.
Okay, so pinpointing problems is key.
Now, something else that sounds really valuable, alerts.
You said these are automated warnings.
How do they help?
Oh, this is where Bezel becomes more than just a viewer.
It's like your digital lookout.
Alerts are your automatic notification system.
Bezel lets you set warnings for, well, all the critical stuff.
CPU usage too high, memory running low, disk getting full, server overheating.
Stuff you definitely want to know about before it causes chaos.
Absolutely.
It sends you a notification before you might even notice a problem yourself.
I remember once running this small data job and the disk just filled up silently,
crashed
everything.
If I'd had an alert saying, hey, disk at 90%, I could have cleaned it up, added
space, avoided
the whole mess.
Right.
It turns a potential disaster into just a quick task.
It's like the low tire pressure light in your car versus getting a flat on the
highway.
Proactive versus reactive.
I wouldn't agree more.
Proactive is always better.
Yeah.
Now what about if you're not working alone, maybe a small team or you want to share
access
to a personal project, Bezel has multi-user features.
It does and they're quite flexible.
Some people can manage their own sets of systems, and if you're an admin, you can
share access
to specific systems with other users.
Okay.
Useful for collaboration.
It also mentions OAuth or IDC support.
Sounds technical.
It's just a secure way to log in using accounts you probably already have, like
Google or
GitHub.
Ah.
Single sign-on.
Makes life easier.
Yeah.
It streamlines login and uses those trusted providers for security.
So you could easily let a friend help troubleshoot one project without sharing your
main server
password, for instance, or bring a collaborator onto an open source project
securely.
Nice.
You can even turn off regular password logins entirely if you want maximum security
that
way.
It really makes managing access flexible and safe.
That level of flexibility is great.
In thinking about safety nets, automatic backups, how important are those, and what
does Bezel
offer?
Hugely important.
It's your ultimate safety net against data loss.
Even with great monitoring, things can happen.
But backups mean your monitoring data and Bezel's configuration gets saved
regularly.
Where does it save them?
You can save them right to the local disk, or even better, to S3-compatible cloud
storage.
Think AWS S3, Backblaze B2, things like that.
It gives you off-site resilience.
Good options.
And you mentioned something for more advanced users, a REST API.
Yes.
This is really neat for tinkerers or developers.
It means you can pull the data Bezel collects out of Bezel and use it in your own
scripts.
Maybe build custom dashboards, integrate it into other tools.
So you're not locked into just using the Bezel interface.
Exactly.
You get programmatic access to all that valuable metric data.
It raises the question of control, right?
How much do you want?
Bezel offers a good amount there.
Okay.
So we have all these features.
How does Bezel actually do all this?
What's the structure?
You mentioned a hub and an agent.
Yeah.
It's a pretty smart, simple architecture, two main parts.
The hub is basically the central brain.
It's the web application you log into.
It's built on something called PocketBase, which is itself a really cool,
lightweight
backend.
Fits the Bezel philosophy perfectly.
So the hub is where you see everything, the dashboard.
Right.
That's your command center.
Then for each machine you want to monitor, you install a small piece of software
called
the agent.
The agent's job is just to sit on that machine, collect all the metrics, CPU,
memory, disk,
all that stuff.
And send it back to the hub.
Yep.
And send the data securely back to the hub using a minimal SSH server connection.
Keeps things safe and efficient.
This hub and agent model makes it really easy to manage lots of systems from one
place.
Got it.
Central hub, multiple agents reporting in.
Makes sense.
Yeah.
So what specific metrics are those agents collecting for us?
What can we actually see?
It's pretty comprehensive.
You get CPU usage, of course, for the whole system.
And for those individual Docker or Podman containers we talked about.
Okay.
Then memory usage, your system's RAM, swap space if you use it, even specialized
stuff
like ZFS, ARC for more advanced setups, again, tracked for the host and containers.
Vital for seeing if you're running out of thinking space.
Exactly.
Then disk usage, how full are your drives?
And also disk IO, how busy are they reading and writing data helps spot storage
bottlenecks.
Right.
Network usage is in there too, data going in and out, again, for the host and
containers.
Good for spotting unusual traffic or bandwidth limits.
And physical health.
Yep.
Temperature from system sensors, crucial for preventing overheating.
And for those using graphics cards for heavy lifting like AI or rendering, there's
specialized
GPU monitoring.
Oh, interesting.
What does that show?
Usage percentage, temperature, even power draw for Nvidia and AMD cards.
You need a specific binary agent for that, but it's powerful if you need it.
Wow.
That covers a lot of ground.
It does.
Each metric tells a piece of the story about your server's health and performance,
helps
you diagnose things fast.
And it sounds like you get this detailed view without it being overly complicated
to set
up, which is the real win here.
You mentioned getting started is easy.
Where do people go?
Yeah, they've put effort into this.
The quick start guide and all the documentation are on their website, bezel.dev, B-E-S-E-E-L.dev.
They honestly claim you can be up and running in a few minutes.
That's a bold claim for server software, but great if true.
It seems pretty streamlined, and importantly, you're not left stranded if you hit a
problem
or have questions.
Support options.
Yeah, the usual places for open source projects.
GitHub issues are there for bug reports or suggesting new features.
And for general chat, asking for help or sharing ideas, they have GitHub
discussions and also
a matrix chat room.
Matrix.
Yeah, the address is hashtag bezel.matrix.org.
So there's a community there to help out.
You're not figuring it all out in isolation.
That community support is often just as important as the tool itself.
So wrapping this up, what's the big takeaway for you, our listener?
Bezel really seems like this elegant answer to making server monitoring less scary.
It gives you those vital insights without bearing you in complexity, turns the
daunting
into something doable, seems perfect if you're just starting running your own
projects or
managing things for a small group.
Absolutely.
And that leads to maybe a final thought to chew on.
Think about the shift in mindset that happens when you go from just reacting to
server problems
to actually understanding how your systems work day to day, getting those quick,
actionable
insights, seeing performance trends, getting alerted before disaster strikes.
It doesn't just save you time and stress.
It frees you up mentally.
Exactly.
It frees up your energy to focus on what you actually want to do, build, create,
learn,
instead of constantly worrying if things are about to break.
That knowledge, that visibility doesn't just help manage servers, it kind of empowers
your whole digital creation process.
It lets you be more ambitious, maybe.
That's a powerful idea.
Understanding empowers creation.
Love it.
And we absolutely have to thank safeserver.de one more time for making this deep
dive possible.
They're dedicated to supporting your digital journey, providing hosting for great
tools
like Bezel.
Thank you so much for joining us on the Deep Dive.
Thank you so much for joining us on the Deep Dive.