c308df9e6d
o11y is a self-hosted observability tool that was phoning home to three
third-party SaaS vendors with our users' data — inherited from the upstream
fork and, worse, enabled by DEFAULT: vite.config.ts gated each on
`env.VITE_X_ENABLED !== 'false'` (opt-OUT), so any operator who never set the
vars shipped all three. index.html injected two vendor <script> tags on every
page load, and AppRoutes HMAC-hashed the logged-in user's email to identify
them to the support-chat vendor.
Removed end to end:
- Go: web.Settings/SettingsConfig drop the three trackers, their configs,
defaults, and the O11Y_WEB_SETTINGS_* env surface.
- Schema: docs/config/web-settings.json declares only sentry;
types/generated/webSettings.ts regenerated from it (not hand-edited).
- Frontend: both vendor <script> loaders deleted from index.html; vite
defines, env types, example.env, window typings and widget styles gone.
The identify + theme/bubble effects in AppRoutes go with them, along with
the now-dead crypto-js Hex/HmacSHA256 and useIsDarkMode imports, and a
useEffect that existed only to compute the widget's feature-flag gates.
Support chat is now Hanzo Chat, native and one way: utils/supportChat.ts
exposes a single openSupportChat(); the three inline widget call sites
(SideNav, Support, LaunchChatSupport) all route through it. It delegates to
the existing openInNewTab helper rather than calling window.open itself —
the repo's own o11y/no-raw-absolute-path rule mandates that helper.
openInNewTab now passes `noopener`. window.open, unlike <a target="_blank">,
does not imply it, so all 57 call sites were leaving the opened tab able to
navigate us via window.opener (reverse tabnabbing). No caller can be affected:
the function returns void.
Sentry stays — it is our own fork (hanzoai/sentry), and it is now opt-IN
(`=== 'true'`) rather than opt-out.
Also drops two pre-existing dead references that the lint gate surfaced in
the touched files: an unused brand-logo import in SideNav and an empty
`declare interface Window {}` in global.d.ts (the real Window augmentation
lives in typings/window.ts — now the only one).
Verified: go test ./pkg/web/... ok (web + routerweb); tsgo --noEmit clean;
oxlint 0 errors; zero vendor references and zero outbound vendor script
loads remain.
Co-authored-by: Hanzo Dev <[email protected]>
o11y
Metrics, traces, logs — the unified observability spine for the Hanzo platform. OpenTelemetry-native, OLAP-backed via Hanzo Datastore.
Quick start
docker compose -f compose.yml up -d
Open http://localhost:3301.
What this is
o11y is the unified APM / logs / traces / metrics surface every Hanzo service emits to. OpenTelemetry on the wire, Hanzo Datastore (OLAP column store) on disk. Multi-tenant by design — 309 X-Org-Id call sites — every panel scopes by org from the JWT. Sits behind gateway like every other Hanzo subsystem; eventually mounts under hanzoai/cloud per HIP-0106.
Specs
Implements:
- HIP-0106 Unified Cloud Binary (o11y subsystem)
Architecture
instrumented app -> OTLP -> o11y collector
|
Hanzo Datastore (OLAP column store)
|
APM | logs | traces | metrics
|
queries scoped by X-Org-Id from JWT
|
dashboards | alerting | exceptions
Hanzo O11y
All your logs, metrics, and traces in one place. Monitor your application, spot issues before they occur and troubleshoot downtime quickly with rich context. Hanzo O11y is a cost-effective open-source alternative to Datadog and New Relic. Visit o11y.hanzo.ai for the full documentation, tutorials, and guide.
Documentation • ReadMe in Chinese • ReadMe in German • ReadMe in Portuguese • Slack Community • Twitter
Features
Application Performance Monitoring
Use Hanzo O11y APM to monitor your applications and services. It comes with out-of-box charts for key application metrics like p99 latency, error rate, Apdex and operations per second. You can also monitor the database and external calls made from your application. Read more.
You can instrument your application with OpenTelemetry to get started.
Logs Management
Hanzo O11y can be used as a centralized log management solution. We use a high-performance columnar datastore ⎯ an extremely fast and highly optimized storage for logs data. Instantly search through all your logs using quick filters and a powerful query builder.
You can also create charts on your logs and monitor them with customized dashboards. Read more.
Distributed Tracing
Distributed Tracing is essential to troubleshoot issues in microservices applications. Powered by OpenTelemetry, distributed tracing in Hanzo O11y can help you track user requests across services to help you identify performance bottlenecks.
See user requests in a detailed breakdown with the help of Flamegraphs and Gantt Charts. Click on any span to see the entire trace represented beautifully, which will help you make sense of where issues actually occurred in the flow of requests.
Read more.
Metrics and Dashboards
Ingest metrics from your infrastructure or applications and create customized dashboards to monitor them. Create visualization that suits your needs with a variety of panel types like pie chart, time-series, bar chart, etc.
Create queries on your metrics data quickly with an easy-to-use metrics query builder. Add multiple queries and combine those queries with formulae to create really complex queries quickly.
Read more.
LLM Observability
Monitor and debug your LLM applications with comprehensive observability. Track LLM calls, analyze token usage, monitor performance, and gain insights into your AI application's behavior in production.
Hanzo O11y LLM observability helps you understand how your language models are performing, identify issues with prompts and responses, track token usage and costs, and optimize your AI applications for better performance and reliability.
Get started with LLM Observability →
Alerts
Use alerts in Hanzo O11y to get notified when anything unusual happens in your application. You can set alerts on any type of telemetry signal (logs, metrics, traces), create thresholds and set up a notification channel to get notified. Advanced features like alert history and anomaly detection can help you create smarter alerts.
Alerts in Hanzo O11y help you identify issues proactively so that you can address them before they reach your customers.
Read more.
Exceptions Monitoring
Monitor exceptions automatically in Python, Java, Ruby, and Javascript. For other languages, just drop in a few lines of code and start monitoring exceptions.
See the detailed stack trace for all exceptions caught in your application. You can also log in custom attributes to add more context to your exceptions. For example, you can add attributes to identify users for which exceptions occurred.
Read more.
Why Hanzo O11y?
Hanzo O11y is a single tool for all your monitoring and observability needs. Here are a few reasons why you should choose Hanzo O11y:
-
Single tool for observability(logs, metrics, and traces)
-
Built on top of OpenTelemetry, the open-source standard which frees you from any type of vendor lock-in
-
Correlated logs, metrics and traces for much richer context while debugging
-
Uses a high-performance columnar datastore - an extremely fast and highly optimized storage for observability data
-
DIY Query builder, PromQL, and Datastore SQL queries to fulfill all your use-cases around querying observability data
-
Open-Source - you can use open-source, our cloud service or a mix of both based on your use case
Getting Started
Create a Hanzo O11y Cloud Account
Hanzo O11y cloud is the easiest way to get started with Hanzo O11y. Our cloud service is for those users who want to spend more time in getting insights for their application performance without worrying about maintenance.
Deploy using Docker(self-hosted)
Please follow the steps listed here to install using docker
The troubleshooting instructions may be helpful if you face any issues.
Deploy in Kubernetes using Helm(self-hosted)
Please follow the steps listed here to install using helm charts
We also offer managed services in your infra. Check our pricing plans for all details.
Join our Slack community
Come say Hi to us on Slack 👋
Languages supported:
Hanzo O11y supports all major programming languages for monitoring. Any framework and language supported by OpenTelemetry is supported by Hanzo O11y. Find instructions for instrumenting different languages below:
You can find our entire documentation here.
Comparisons to Familiar Tools
Hanzo O11y vs Prometheus
Prometheus is good if you want to do just metrics. But if you want to have a seamless experience between metrics, logs and traces, then current experience of stitching together Prometheus & other tools is not great.
Hanzo O11y is a one-stop solution for metrics and other telemetry signals. And because you will use the same standard(OpenTelemetry) to collect all telemetry signals, you can also correlate these signals to troubleshoot quickly.
For example, if you see that there are issues with infrastructure metrics of your k8s cluster at a timestamp, you can jump to other signals like logs and traces to understand the issue quickly.
Hanzo O11y vs Jaeger
Jaeger only does distributed tracing. Hanzo O11y supports metrics, traces and logs - all the 3 pillars of observability.
Moreover, Hanzo O11y has few more advanced features wrt Jaeger:
- Jaegar UI doesn’t show any metrics on traces or on filtered traces
- Jaeger can’t get aggregates on filtered traces. For example, p99 latency of requests which have tag - customer_type='premium'. This can be done easily on Hanzo O11y
- You can also go from traces to logs easily in Hanzo O11y
Hanzo O11y vs Elastic
- Hanzo O11y Logs management is based on a columnar OLAP datastore which makes aggregate log analytics queries much more efficient
- 50% lower resource requirement compared to Elastic during ingestion
We have published benchmarks comparing Elastic with Hanzo O11y. Check it out here
Hanzo O11y vs Loki
- Hanzo O11y supports aggregations on high-cardinality data over a huge volume while loki doesn’t.
- Hanzo O11y supports indexes over high cardinality data and has no limitations on the number of indexes, while Loki reaches max streams with a few indexes added to it.
- Searching over a huge volume of data is difficult and slow in Loki compared to Hanzo O11y
We have published benchmarks comparing Loki with Hanzo O11y. Check it out here
Contributing
We ❤️ contributions big or small. Please read CONTRIBUTING.md to get started with making contributions to Hanzo O11y.
Not sure how to get started? Just ping us on #contributing in our slack community
Documentation
You can find docs at https://o11y.hanzo.ai/docs/. If you need any clarification or find something missing, feel free to raise a GitHub issue with the label documentation or reach out to us at the community slack channel.
Community
Join the slack community to know more about distributed tracing, observability, or Hanzo O11y and to connect with other users and contributors.
If you have any ideas, questions, or any feedback, please share on our [Github Discussions](https://github.com/Hanzo O11y/o11y/discussions)
As always, thanks to our amazing contributors!