2026.10.23
HyperDX: A Unified Place for Logs, Traces, and Session Replays
HyperDX is an open-source observability platform built around ClickHouse and OpenTelemetry. This guide covers its capabilities, target users, deployment model, and adoption constraints.
What it is for
HyperDX is an open-source observability platform designed to help engineers investigate production problems through a shared view of logs, metrics, traces, errors, and session replays. It is built around ClickHouse and supports OpenTelemetry as a major ingestion path. Teams already using ClickHouse, or those seeking one place to correlate several telemetry types, may find it worth evaluating.
The project is also presented as a component of ClickStack. A full deployment can include ClickHouse, HyperDX, the OpenTelemetry Collector, and MongoDB. Existing ClickHouse users can choose a deployment option documented by the project. For testing, the README recommends at least 4 GB of RAM and two CPU cores. That is a testing recommendation, not a production capacity plan.
Image source: hyperdx. Published by the project; the interface may vary by version.
Capabilities and suitable users
HyperDX can work over an existing ClickHouse schema. Its documented capabilities include full-text and property search, optional SQL, native JSON string querying, anomaly trend analysis, alerts, live log and trace views, and dashboards for high-cardinality events. It also covers application performance monitoring from HTTP requests to database queries.
Development teams, platform engineers, and SREs who need to connect evidence across logs, traces, and frontend sessions may benefit from this unified workflow. The project lists Browser, Node.js, and Python SDKs, and documents OpenTelemetry support for ecosystems including Kubernetes, JavaScript, Python, Java, Go, Ruby, PHP, .NET, Elixir, and Rust. Actual integration coverage, field mapping, and collection completeness should still be verified for each target system.
For terminal-oriented workflows, HyperDX also provides a CLI and interactive TUI. The project describes support for search, live tailing, and trace waterfalls, along with dashboard tiles and ad-hoc chart output. The official 0.7.0 CLI release added a stacked line or stacked area time-series display type.
Constraints and deployment considerations
HyperDX does not create useful telemetry by itself. Applications still need SDK instrumentation or OpenTelemetry pipelines that send data to the deployment. Without reliable collection, consistent fields, and a retention policy, a unified interface cannot replace telemetry governance. When using an existing ClickHouse instance, teams should review its schema, volume, query patterns, and retention requirements before adoption.
The complete stack also introduces several operational components. Define ownership, backup procedures, upgrade paths, and network boundaries for ClickHouse, the Collector, MongoDB, and HyperDX. The README notes that firewall deployments need access handled for UI, API, and OpenTelemetry Collector ports. Production exposure, authentication, reverse-proxy configuration, and high availability should be validated against the official deployment documentation and the target environment. The supplied sources do not establish production performance, capacity, or security guarantees.
Open-source deployments collect anonymized usage data according to the README, with an opt-out through USAGE_STATS_ENABLED=false. Organizations should include this in privacy, compliance, and network reviews. The repository uses the MIT License, but dependencies, retention, and sensitive data in session replays still require separate review.
Overall, HyperDX is most relevant to teams that want unified telemetry search, already use or are willing to adopt ClickHouse, and can operate the ingestion and supporting services. Production suitability should be established with representative data and tests covering collection, query behavior, cost, and operational boundaries—not from the feature list alone.
Sources
Sources checked: 2026-10-23