DolphinDB

DolphinDB is a real-time analytics and stream-processing platform built around a high-performance time-series database. It is best suited to teams handling demanding, mixed workloads that need streaming, batch analysis, and distributed deployment in one system. The trade-off is a broad platform to operate, with a free edition sized for small deployments.

Overview

DolphinDB combines storage engines for time-series databases, OLAP, and in-memory OLTP. That makes it a candidate for organizations that would otherwise assemble separate systems for market or sensor series, analytical warehousing, and low-latency transactional work. Cloud, on-premise, and hybrid deployment are supported, with horizontal cluster scaling when a small installation is no longer sufficient.

The breadth is worthwhile when several of those workloads must share data and operational controls. A team needing only a narrowly focused time-series store may find the extra engines, languages, and integrations unnecessary.

Key features

More than 1,500 built-in analytical functions and more than 10 streaming engines give DolphinDB substantial scope for real-time analysis. Batch and stream processing can use the same metric expression, helping teams avoid maintaining separate definitions for historical and live calculations.

Programming is flexible: imperative, vectorized, functional, and SQL styles are supported. Query choices include DolphinDB SQL, ANSI SQL/SQL-92, and the DolphinDB scripting language. Data can be ingested from text, binary, JSON, HDF5, Parquet, Arrow, ODBC, APIs, Kafka, MQTT, ZeroMQ, MySQL, MongoDB, HBase, and kdb+. This is a strong fit for heterogeneous pipelines, although each connector adds another component to govern.

Distributed features include horizontal scaling, load balancing, fault tolerance, distributed transactions, and high availability across data nodes, controllers, streaming services, and API clients. Retention policies and SQL support are included. Security controls cover users and groups, multiple privilege types, function views, RSA encryption, and HTTPS.

The DolphinDB VS Code extension supports writing and executing scripts, code completion, multi-user connections, version control, and debugging. The Python API documents Windows amd64, Linux x86_64 and aarch64, and Mac x86_64 and arm64 environments. Kafka publishing and subscription support JSON, DolphinDB, and string formats; on ARM Linux, SASL authentication and zstd compatibility are absent, a material constraint for those deployments.

Additional components target particular workloads: Swordfish embeds DolphinDB, Shark provides a GPU version, Starfish supports quantitative research workflows, and Octopus is a complex event-processing engine.

Pricing

DolphinDB Community Edition

The Community Edition costs free on a perpetual license. It is limited to up to two nodes, with two CPU cores and 8 GB of RAM per node. That is useful for evaluation, development, and compact production workloads, but the node and per-node resource caps make it unsuitable as a no-cost path to a larger highly distributed deployment.

DolphinDB Enterprise Trial

The Enterprise Trial also costs free and supports high-availability clusters with expert technical support. It is the more appropriate starting point for an organization assessing resilient multi-node operation, while the Community Edition is better for a constrained, self-directed setup. A free trial is available.

Platforms

DolphinDB is offered through APIs, extensions, Linux, self-hosted deployment, the web, and Windows. Both cloud and on-premise deployment options are supported, as well as hybrid arrangements. The documented Python environments also include macOS architectures, making the API practical across common development machines.

Who it's for

Finance, energy, power, transportation, industrial manufacturing, and scientific-research teams are the clearest fits. Choose DolphinDB when one platform must combine time-series data, analytical and transactional engines, streaming calculations, broad ingestion, and distributed availability. It is less compelling for a small project that needs one simple database and has no use for cluster operations or multiple programming models.

Pros and cons

• Pro: Time-series, OLAP, and in-memory OLTP engines cover materially different workloads in one platform.
• Pro: Shared expressions for batch and streaming analysis can reduce duplicated metric logic.
• Pro: High availability, fault tolerance, load balancing, and distributed transactions support serious clustered deployments.
• Pro: The integration surface is unusually broad, including Kafka, MQTT, files, databases, and kdb+.
• Con: Community Edition capacity stops at two nodes and 2 CPU cores plus 8 GB RAM per node, limiting free production scale.
• Con: The many engines, languages, connectors, and deployment modes demand more operational discipline than a focused database.
• Con: ARM Linux Kafka deployments lack SASL authentication and zstd compatibility.

Alternatives

For a conventional analytical database with published pay-as-you-go sizes, consider Exasol, whose XS plan provides 8 vCPUs and 64 GB of memory for $3.30 a month and whose S plan doubles those resources for 6.60 USD.

InfluxDB is a better fit when a time-series-focused product is preferable; its InfluxDB 3 Core plan is single-node, while Cloud Serverless charges for what you use.

Choose QuestDB for an Apache 2.0, self-managed single instance with community support. TDengine suits teams wanting an AGPL open-source time-series database with clustering in its OSS edition.

Timeplus is an alternative for lightweight streaming SQL through its Apache 2.0 Proton single binary. Apache Druid offers an open-source analytics database for self-hosting, while DuckDB UI is aimed at local SQL notebooks with optional MotherDuck connections. Teradata VantageCloud is the enterprise-oriented alternative.

For broader category browsing, see OLAP Software, Multimodel Databases, and Time-Series Databases.

Verdict

Choose DolphinDB if your organization needs a distributed platform that unifies time-series, analytical, transactional, and streaming workloads, especially in finance or industrial data operations. Its main reason to look elsewhere is complexity—or the Community Edition’s tight two-node resource ceiling when you need a larger free deployment.