- Home
- Disk Benchmark Software
- Storage Benchmark Kit
Storage Benchmark Kit
What it is
Storage Benchmark Kit (SBK) is a free Java framework for measuring storage throughput and latency with a shared workload engine. Drivers cover object stores, message systems, databases, file systems, caches, and local queues. The harness controls concurrency, duration, rate, payloads, timestamps, and reporting. Results can be sent to console, CSV, a local web console, Prometheus, or gRPC. PrometheusLogger exposes metrics, and SBK Dashboard supports Grafana dashboards with persistent time-series history. SBK-GEM runs distributed benchmarks by copying and launching SBK on remote hosts over SSH, then aggregating measurements. YAML tools load arguments for single-node and distributed runs. SBK requires JDK 25; its wrapper can download a pinned OpenJDK release and verify its SHA-256 checksum. Automatic installation supports Linux and macOS on x64 or AArch64, plus Windows x64. It is licensed under Apache License 2.0.
For and against
For
- Measures throughput and latency
- Supports many storage backend categories
- Configurable concurrency and workload parameters
- Prometheus and Grafana support
- Free Apache License 2.0 project
Against
- Requires JDK 25
- Local web console lacks authentication and TLS
- Some drivers are excluded from aggregate build
- Distributed runs require SSH access
The verdict
SBK provides a configurable benchmark engine with local, metrics, and distributed output options. Users should protect its plain-HTTP local console and account for the JDK requirement and excluded drivers.
Who it suits
SBK suits engineers who need repeatable storage benchmarks, configurable workloads, metrics retention, or distributed execution. Documentation also covers building drivers and running benchmark workflows.
Plans and prices
Apache License 2.0
Open-source license · no paid plans listed
Read 9 Oct 2026 · github.com
On the record
- Backend coverage
- Its drivers cover object stores, message systems, databases, file systems, caches, and local queues.kmgowda.github.io · read 9 Oct 2026
- Configuration
- SBK-YAL loads single-node SBK arguments from YML, while SBK-GEM-YAL loads distributed SBK-GEM arguments from YML.kmgowda.github.io · read 9 Oct 2026
- Distributed execution
- SBK-GEM copies and launches SBK on remote hosts over SSH and runs an embedded SBM instance to aggregate measurements.kmgowda.github.io · read 9 Oct 2026
- Integrations
- PrometheusLogger exposes metrics for Prometheus, and SBK Dashboard supports Grafana dashboards and persistent time-series history.kmgowda.github.io · read 9 Oct 2026
- Intended users
- The documentation provides workflows for running benchmarks, building storage drivers, retaining metrics, and executing distributed benchmarks.kmgowda.github.io · read 9 Oct 2026
- License
- SBK is licensed under the Apache License 2.0.github.com · read 9 Oct 2026
- Metrics
- SBK includes console, CSV, local web console, Prometheus, and gRPC output options.kmgowda.github.io · read 9 Oct 2026
- Notable availability limit
- ChromaDB, HaloDB, and Ignite are in the source tree but are not enabled in the aggregate build.kmgowda.github.io · read 9 Oct 2026
- Operating systems
- Automatic JDK installation supports Linux x64/AArch64, macOS x64/AArch64, and Windows x64.kmgowda.github.io · read 9 Oct 2026
- Purpose
- SBK is a Java framework for measuring storage system throughput and latency with a common workload engine.kmgowda.github.io · read 9 Oct 2026
- Runtime requirement
- SBK requires JDK 25; its wrapper can download a pinned OpenJDK 25 release and verify its SHA-256 checksum.kmgowda.github.io · read 9 Oct 2026
- Security
- The Local Web Console uses plain HTTP without authentication or TLS, and the project advises exposing it only on a trusted benchmark network or through an SSH tunnel.kmgowda.github.io · read 9 Oct 2026
- Support
- The project directs bug reports and feature requests to GitHub Issues and usage or design questions to GitHub Discussions.kmgowda.github.io · read 9 Oct 2026
- Workload controls
- The harness controls concurrency, duration, rate, payloads, timestamps, and reporting.kmgowda.github.io · read 9 Oct 2026
