GreptimeDB Dashboard has no built-in authentication: what nineteen title matches imply
GreptimeDB Dashboard has no built-in authentication: what nineteen title matches imply GreptimeDB's open-source Dashboard is a browser interface for querying metrics, logs, and traces. The documentation states that the
GreptimeDB Dashboard has no built-in authentication: what nineteen title matches imply
GreptimeDB's open-source Dashboard is a browser interface for querying metrics, logs, and traces. The documentation states that the open-source Dashboard does not include built-in authentication. Operators can place it behind a reverse proxy or internal authentication system. That design fact is more useful than the raw title count.
On November 17, 2026, a count-only ZoomEye measurement used sub_type=all, page 1, and a page size of 1. title="GreptimeDB" returned 19 matches. The app="GreptimeDB" probe failed with an SSL EOF transport error, so it is not a zero and its count is unavailable. The title result is a successful observation. The failed probe is recorded separately and should not be used to infer absence.
What the counts can support
GreptimeDB's HTTP endpoint documentation explains two servers. The main HTTP server on port 4000 serves operational and admin paths, including /dashboard, /config, /metrics, and debug routes. The optional public API server on port 4006 serves only /v1 APIs and /dashboard and is disabled by default. The documentation warns that route filtering does not provide authentication or transport security. With no user provider configured, requests can be accepted without authentication.
That warning turns a small title count into a configuration checklist. Is the Dashboard exposed through port 4000 or 4006? Is the main server reachable only by trusted operators? Is a user provider configured? Is TLS terminated at GreptimeDB or at a trusted proxy? Does the proxy authenticate users before forwarding /dashboard? Does it also protect API paths that can execute SQL or PromQL?
The dashboard is not just a read-only visualisation. It can send queries to the database. The documentation describes a general query entry and views for metrics, logs, and traces. A public Dashboard with no authentication can therefore disclose or alter data depending on the configured user and database privileges. The exact impact depends on the service configuration, which a ZoomEye title cannot show.
For owned assets, record the GreptimeDB version, main and public listener addresses, dashboard enablement, reverse proxy rules, TLS mode, authentication provider, configured databases, and query privileges. Confirm the running configuration after upgrades. The release notes for version 1.3.1 describe database authorization and credential exposure fixes, so version drift is relevant to the overall review even though it is separate from this exposure measurement.
A reverse proxy is a valid control if it is configured as a security boundary. Basic authentication, SSO, source-network restrictions, request-size limits, and TLS should be tested from an external network. A proxy that only rewrites the path and forwards all headers does not provide authentication. A static Dashboard page may load successfully even when its API calls fail, so test the query path as well as the landing page.
Verification and next steps
Passive attribution can use certificate subjects, approved domains, and page titles. Do not attempt to execute queries against unknown instances. If a title match belongs to the organisation, use configuration and an approved test account to verify the controls. If the Dashboard is intentionally public, document the data classification and the compensating authentication layer.
The app probe's SSL failure should be visible in the evidence record. It may be a transient network or service condition, but it is not a zero count. A later measurement can compare the title result and a successful app query under the same scope. Do not silently replace the failed row with a guess.
GreptimeDB is designed for observability data, which often contains service names, host details, request patterns, and operational metadata. The Dashboard is useful for exploring that data. Nineteen title matches justify a focused review of where the Dashboard lives and who can reach it. The central control is simple: the open-source Dashboard needs an external authentication and network boundary, and the query API behind it needs the same review.
References
Originally published by Dev.to Security. Aggregated on AIWithGhost for educational purposes β full credit and traffic to the original publisher.