GreptimeDB

GreptimeDB is an open-source observability database for storing and querying metrics, logs, and traces in a columnar engine with a shared SQL layer. It targets append-heavy, time-indexed workloads such as observability and IoT data, rather than general-purpose transactional work. Users can query signal tables with SQL and metrics with PromQL. Ingestion connections include MySQL, PostgreSQL, gRPC, Prometheus, OpenTelemetry, and InfluxDB line protocol, while documented integrations include Grafana, Kafka, Apache Flink, Apache Spark, Telegraf, EMQX, and Fluent Bit. It can run as a standalone database or a horizontally scalable cluster; downloads include Helm charts for Kubernetes, Docker images, and binaries for macOS, Linux, Windows, and Android ARM64. Its feature list includes object storage and indexes for high-cardinality data. The open-source edition lists password authentication, user-level read/write controls, and TLS/SSL. A free plan is available, and the fully managed Enterprise plan starts at $290/mo. Enterprise adds features including fine-grained access controls, audit logging, and encryption at rest.

GreptimeDB combines several observability signals and ingestion paths, with standalone and cluster deployment options. Its free open-source plan contrasts with managed Enterprise pricing that starts at $290/mo.

GreptimeDB is suited to teams handling time-indexed observability or IoT data who want SQL and PromQL querying. It is not positioned for general-purpose transactional database workloads.

Overview

GreptimeDB is aimed at teams that need a database for observability or IoT streams, not a general-purpose transactional store. Its appeal is the combination of multiple signal types, query options, and deployment choices; the trade-off is that its strengths are specialized around time-indexed, append-heavy data.

Metrics, logs, and traces share a columnar engine and SQL layer, with PromQL for metrics. Data can be ingested through familiar database connections and observability protocols, while deployment can range from a standalone instance to a horizontally scalable cluster.

It fits both Time-Series Databases and Columnar Databases, particularly for readers evaluating an observability-focused system rather than a transactional database.

Key features

• Multiple signals, shared querying: Metrics, logs, and traces use the same engine, and SQL works across signal tables. PromQL provides a metrics-specific option, with Log Query also supported. This can reduce the need to approach each signal through an entirely separate data model.
• Flexible ingestion: Endpoints include Prometheus Remote Write, OpenTelemetry OTLP/HTTP, Loki Push API, Elasticsearch Bulk API, InfluxDB Line Protocol, MySQL, PostgreSQL, gRPC, and HTTP. Documented integrations include Prometheus, Grafana, Kafka, Apache Flink, Apache Spark, Telegraf, EMQX, and Fluent Bit.
• Scaling and storage: GreptimeDB supports standalone operation as well as horizontally scalable clusters. Object storage support and inverted, skipping, and full-text indexes are listed for high-cardinality data, a useful combination for varied observability workloads.
• Security varies by edition: Open source includes username/password authentication, user-level read/write access control, and TLS/SSL connections. Enterprise and Cloud add RBAC, LDAP, encryption at rest, and audit logging; Enterprise also specifies encryption in transit.
• Enterprise capabilities: The Enterprise edition adds read replicas, automated region operations, SQL-native triggers, and an Iceberg REST catalog. These features broaden the product beyond its open-source baseline, but make plan selection important for teams needing advanced operations or governance.

Pricing

GreptimeDB uses a freemium model. The open-source plan is free and self-hosted, with community support through Slack and GitHub. It gives technical teams a way to run the database without a subscription, but Enterprise security features, one-on-one technical support, and SLA-backed response times require a paid Enterprise option.

GreptimeDB Enterprise Fully-Managed: $290 a month, billed From $290/month. It includes unlimited data storage and retention, isolated resource control and network, and a customizable, fully managed deployment topology. This is the clearest fit for organizations that want managed operations and no stated storage or retention cap; its starting price makes it a materially different commitment from the free self-hosted edition.

GreptimeDB Enterprise Bring Your Own Cloud: custom pricing. It includes deployment in your cloud, a custom SLA, and advanced technical support. This suits teams that need to keep deployment in their own cloud and want tailored service terms, though the custom price requires a direct commercial decision rather than a published monthly comparison.

Platforms

GreptimeDB supports Android, API access, Linux, macOS, self-hosting, web, and Windows. Download options include Kubernetes Helm charts, Docker images, and binaries; listed binaries cover macOS, Linux, Windows, and Android ARM64. This breadth gives teams several deployment routes, though platform availability should not be confused with a turnkey desktop application.

Who it's for

GreptimeDB is best suited to engineering and data teams collecting high-volume, time-indexed observability or IoT data and wanting metrics, logs, and traces in one system. Its protocol support and integrations help it sit alongside existing telemetry pipelines, while the choice of standalone or cluster deployment suits projects with different scaling needs.

It is not the natural choice for a general-purpose transactional workload. Teams that prioritize a mature managed service with clearly published per-use costs, or that do not need a specialized observability database, should compare alternatives before committing.

Pros and cons

• Pro: One columnar engine covers metrics, logs, and traces, with SQL across signal tables and PromQL for metrics; this can simplify querying across observability data.
• Pro: Ingestion options span common database protocols and telemetry APIs, making it relevant to varied collection pipelines.
• Pro: Open-source self-hosting is free, and deployment can scale from standalone to cluster.
• Pro: Enterprise offers managed and bring-your-own-cloud choices, alongside advanced security and SLA-backed support.
• Con: The design targets append-heavy, time-indexed workloads, so it is a poor fit for general transactional database use.
• Con: Advanced security, dedicated technical support, and enterprise operational capabilities are gated behind paid Enterprise plans.
• Con: Bring Your Own Cloud uses custom pricing, which does not provide a published price for straightforward budget comparison.

Alternatives

• Couchbase Lite is a paid option without a free plan, with a Couchbase Mobile plan that includes Sync Gateway licensing, peer-to-peer syncing, SQL++, full-text search, and vector search. Consider it when those sync and search capabilities are central rather than GreptimeDB's observability focus.
• Exasol has a free plan and free trial, with XS at $3.30 a month and S at $6.60 a month, both pay-as-you-go. Consider it when a freemium database with those listed compute tiers better matches your needs.
• InfluxDB is freemium, with a free plan and trial; its Cloud Serverless - Data In plan is free on a pay-for-what-you-use basis. Consider it when that serverless usage-based option is a better fit.
• Apache Druid is a free, downloadable, self-hosted open-source analytics database. Consider it for self-hosted analytics when that framing is a closer match.
• CrateDB is freemium; its Shared Free plan provides 2 vCPUs, 2 GiB RAM, and 8 GiB storage at no cost. Consider it when a defined free shared tier with those resource limits suits the workload.
• Amazon Timestream is paid, with no free plan; its on-demand pricing has no minimum fees or upfront commitments and depends on the engine and usage. Consider it when that pay-as-you-go model is preferable.
• Databend offers a free self-hosted Community plan for deployment in your own environment. Consider it when a free self-hosted plan is the priority.
• MotherDuck has a free Lite plan with up to 3 internal active users, 2 service accounts, 10 GB free storage, and 10 hours of Pulse compute per month, limited to Pulse instances with community support. Consider it when those user and compute limits fit.

Verdict

Choose GreptimeDB if your team needs a specialized observability database that brings metrics, logs, and traces together and offers both self-hosted and managed deployment paths. Its strongest case is the combination of broad ingestion, shared querying, and scaling options. Look elsewhere for general transactional workloads, or if Enterprise pricing and features are not justified by your operational needs.