> ## Content Index
> Fetch the complete content index at: https://ditchthesaas.com/llms.txt
> Use this file to discover other available public pages before exploring further.

# Free Datadog Alternative: Self-Hosted Monitoring with Grafana + Loki
- URL: https://ditchthesaas.com/free-datadog-alternative-self-hosted-monitoring-with-grafana-loki/
- Published: 2026-09-12T14:52:30.000Z
- Updated: 2026-09-12T14:52:30.000Z
- Description: Datadog's per-host pricing adds up fast. Here's what it actually costs to run your own monitoring stack instead, and what you give up.
- Author: Jordan Cole
- Tags: Self-Hosted Stack

*Second post in the series on replacing SaaS subscriptions with self-hosted tools. First one covered* [*automation (n8n vs Zapier)*](https://ditchthesaas.com/n8n-vs-zapier-how-i-cut-my-automation-bill-by-self-hosting/) *— this one covers monitoring.*

**Disclosure:** this post contains an affiliate link to Vultr. If you sign up through it, I may earn a commission at no extra cost to you. I only link to providers I actually use.

## Why Datadog bills get out of hand

Datadog's pricing looks reasonable until you actually turn things on. As of 2026:

- **Infrastructure Monitoring**: $15/host/month (annual), $18/host on-demand
- **APM**: $31/host/month, on top of infrastructure monitoring
- **Log Management**: $0.10/GB for indexed logs, plus ingestion fees
- Custom metrics, high-cardinality tags, and extended retention are where bills quietly explode

None of this is outrageous for a well-funded team that wants zero ops overhead. But for a handful of servers, it adds up to real money for something you could run yourself.

## What self-hosted monitoring actually costs

Grafana and Loki are both free and open source — no per-host or per-GB licensing, ever. Your only cost is the server they run on.

- A small VPS is enough to run Grafana + Loki for a handful of hosts — the same size instance that runs n8n in the [previous post](https://ditchthesaas.com/n8n-vs-zapier-how-i-cut-my-automation-bill-by-self-hosting/) can usually handle both
- Grafana Cloud's free tier exists too (10,000 series, 50GB logs, 14-day retention, 3 users) — fine for testing, but production workloads tend to outgrow the series cap fast
- Self-hosting means no series caps, no retention limits, and no per-GB log bill — just your own disk

## The real trade-off

Nobody hands you a fully managed dashboard on day one. You're responsible for keeping Loki's components running, managing storage growth, and upgrading versions yourself. On a small homelab-scale setup this is a few hours of setup and occasional maintenance — not a full-time job — but it's not zero, either. If your team doesn't have anyone willing to own that, Datadog's price starts looking more reasonable.

## Cost comparison

|                      | Datadog                           | Self-hosted Grafana + Loki   |
| -------------------- | --------------------------------- | ---------------------------- |
| Monthly cost         | $15–$27+/host, APM +$31/host      | VPS: $5–20 total             |
| Log costs            | $0.10/GB indexed + ingestion fees | Disk space only              |
| Series/metric limits | Custom metrics billed separately  | No caps                      |
| Maintenance          | Fully managed                     | You run upgrades and storage |

## Getting started

Grafana + Loki run comfortably on the same class of VPS as the rest of this stack. I use [Vultr](https://www.vultr.com/?ref=9922680) for this — a small instance handles Grafana, Loki, and a handful of scrape targets without issue. Start small and only scale up once you actually see memory pressure.

*Next up: Vikunja as a self-hosted Notion/Asana alternative.*