programming Monitoring & Observability — 7 tools curated by the community. Most popular: SigNoz, Honeycomb, Highlight.io. Each tool comes with a review, tags and alternatives to help you choose.
❓ Frequently asked questions about Monitoring & Observability
What is Monitoring & Observability?
Monitoring & Observability groups 7 tools curated by the mySelectas community. Most popular: SigNoz.
What are the best free Monitoring & Observability tools?
mySelectas lists several free options in this category. See in particular: SigNoz, Honeycomb, Highlight.io.
The short version: if your team is paying Datadog or New Relic by the gigabyte and mostly just wants traces, logs and dashboards in one place, SigNoz gives you that for free if you're willing to self-host, or a fraction of the price if you're not.
SigNoz's core bet is that OpenTelemetry has won as the instrumentation standard, so it built the whole platform around ingesting OTel data natively rather than bolting it on — which shows up in cleaner correlation between traces, logs and metrics than most retrofitted tools manage. 31.3k GitHub stars and an MIT-licensed core back up that it's a real…
The short version: Honeycomb is the observability tool built for a specific kind of pain — "it's broken for some users, not all, and I don't know why" — and it solves that faster than most competitors thanks to BubbleUp, at the cost of not having a free self-hosted option.
Honeycomb's differentiator over generic APM dashboards is that BubbleUp turns root-cause analysis into a comparison operation instead of a manual filtering exercise — you select the anomalous slice of traffic and it tells you what it has in common, which genuinely saves the hours engineers otherwise spend eyeballing dashboa…
The short version: Highlight.io is what you'd get if session-replay tools (LogRocket, FullStory) and error/observability tools (Sentry, Datadog) had a baby that's also open-source and self-hostable — genuinely useful if you want both in one place instead of paying two vendors.
The real value is the correlation: a session replay linked directly to the backend error and trace that happened at the same timestamp removes the manual work of cross-referencing a frontend bug report with backend logs. Apache 2.0 licensing and Docker self-hosting give it a credible open-source story (9.3k GitHub stars…
The short version: if you just want to know how many people visit your site and where from — nothing more — GoatCounter is about as simple and privacy-clean as analytics gets, for free.
GoatCounter's whole design philosophy is subtraction: no user profiles, no heatmaps, no session replay, no consent banner requirement in most places because it simply doesn't collect what triggers one. That makes it the right tool for developers, indie makers and small businesses who find Google Analytics both bloated and privacy-uncomfortable, without needing Plausible or Fastly-scale infrastructure either. T…
👥 Target audience
Indie developers | Bloggers | Privacy-conscious small businesses
The short version: the fastest way to see what your AI calls actually cost and to stop depending on a single provider — with one architectural trade-off to understand before adopting it.
Helicone's pitch is pragmatic: point your base URL at their proxy and you instantly get cost tracking, latency metrics and logs for every LLM call — no SDK rewiring. That one-line adoption is its best feature, and combining gateway (routing, fallbacks, one key for 100+ models) with observability means fewer moving parts than assembling OpenRouter plus a logging tool. The trade-offs are structural: a proxy sit…
The short version: one database instead of three for server-health data — the idea is right and the engineering is serious, but the ecosystem around it is still young.
GreptimeDB's premise is the correct diagnosis of a real disease: running Prometheus, Loki and Elasticsearch side by side means three query languages, three retention policies and three on-call runbooks for what is conceptually one dataset. Unifying signals in a columnar store with object storage as primary — so retention costs S3 prices, not SSD prices — is the architecture most observability vendors are converging on, and doin…
The short version: a young but serious tool that automates the expert work of keeping a Postgres database healthy — made by a team that has done exactly that job for a decade.
PostgresAI encodes real, hard-won expertise: the postgres.ai team (Database Lab Engine, postgres-checkup) has done production Postgres work for companies like GitLab and Miro for years, and the 45+ health checks reflect that lineage rather than a thin AI wrapper. The agent-native design — structured JSON diagnostics, MCP server, Claude Code plugin — is a genuine differentiator in 2026; most competitors still assume a hu…
👥 Target audience
Backend & DevOps engineers | Teams without a dedicated DBA
