Table of Contents
| episode | airdate | recorded | status | short_title | full_title | description | summary | guest | guest_title | guest_links | headshot | youtube_url | youtube_id | fireside_player | slug | tags | topics_to_avoid | ||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| 025 | 2026-07-21 | published | The Homelab SLA | The Homelab SLA | The IT Guy Show 025 | Mathias Palmersheim runs his personal homelab against a real SLA. We get into VictoriaMetrics, open source observability, and applying enterprise rigor at home. | Mathias Palmersheim brings enterprise monitoring discipline home and runs his homelab against an actual SLA. | Mathias Palmersheim | Solutions Engineer at VictoriaMetrics |
|
host.png | https://youtu.be/KCljAfOn89Y | KCljAfOn89Y | B1g3qpQH+6d8P41Xq | itguyshow-ep25-homelab-sla |
|
Abstract
Most of us treat our homelabs like a junk drawer. Something breaks, we shrug, we fix it whenever we get around to it. Mathias Palmersheim does not operate that way. He's a Solutions Engineer at VictoriaMetrics, where he spends his days helping companies monitor infrastructure at serious scale, and he's brought that same discipline home. He runs his personal homelab against an actual SLA, the kind of uptime commitment most of us only deal with at work.
Mathias joins me to talk about VictoriaMetrics itself, the time series database built for high cardinality metrics and logs at scale, and why open source observability tooling matters even outside the enterprise. But the real fun is in how he applies that professional rigor to his own gear at home. We get into what it actually takes to hold yourself to an SLA when there's no boss checking in, what breaks the plan, and whether this level of intensity is something the rest of us should aspire to or just admire from a distance.
If you've ever wondered what happens when a monitoring expert points all that expertise at his own home network, this one's for you.
Hook
Most people's homelab uptime commitment is "I'll get to it eventually." Mathias Palmersheim runs his against an actual SLA, and today we find out what that really means.
Agenda
VictoriaMetrics: The Product
- What VictoriaMetrics is and the problem it solves
- Time series data and why high cardinality gets painful fast
- Open source core vs. enterprise and cloud offerings
- Where VictoriaLogs fits in
Mathias's Role and ShiftMon
- What a Solutions Engineer does day to day
- ShiftMon and an open source on-ramp to observability
- Why he built it and who it's for
The Homelab SLA
- What the SLA actually covers: uptime targets, response times
- What's running that justifies the standard
- How he monitors his own monitoring, and what happens when he misses
Professional Rigor Meets Personal Project
- Where useful discipline ends and overkill begins
- Enterprise monitoring lessons that translate home
- Advice for someone starting a homelab who wants to run it seriously
Wrap Up
- Where to find Mathias
- Final thoughts
Keywords
victoriametrics, observability, monitoring, time series, high cardinality, victorialogs, homelab, sla, open source, prometheus, metrics, shiftmon, infrastructure monitoring
Links
- Linktree: https://linktr.ee/itguyeric
- Mathias: https://links.shiftsystems.net
- VictoriaMetrics: https://victoriametrics.com/
Blog Post
Most of us treat our homelabs like a junk drawer. Mathias Palmersheim does not. He does monitoring at serious scale for a living at VictoriaMetrics, and he brought that same discipline home, running his personal lab against an actual SLA.
{{< episode >}}
A few things worth carrying away from this one:
- VictoriaMetrics, briefly. We covered what the time series database is for, why high cardinality gets painful fast, and why open source observability matters even outside the enterprise.
- Holding yourself to an SLA. What it actually takes to run a real uptime commitment at home when there is no boss checking, and what breaks the plan.
- Rigor worth borrowing. Which enterprise monitoring habits translate to a home setup, and where it tips into overkill.
If you've ever wondered what happens when a monitoring expert points all that expertise at his own network, this is your answer.