Welcome: feedback, platform previews, issues, and new workload ideas

[Bob Ward (rgward)]

Welcome to sqlsimtools Discussions.

sqlsimtools is in Public Preview. Discussions are the place for usage questions, platform-testing interest, design feedback, and new workload ideas. Reproducible defects remain in GitHub Issues so they can be tracked through resolution.

Choose the right channel

• Bug or reproducible failure: search existing issues, then open a new issue. Include the tool and version, operating system and architecture, SQL target and version, authentication mode (never the credential), masked command, reproduction steps, and JSON output when available.
• Security vulnerability: do not post it in a discussion or public issue. Follow SECURITY.md and report it to the Microsoft Security Response Center.
• Usage question: start a Q&A discussion.
• New simulator or feature idea: start an Ideas discussion.
• Something you built: use Show and tell.

Platforms and preview access

• Public Preview: sqlsim and sqlpysim on Windows and Ubuntu 24.04/26.04 x64, and sqlsim Studio on Windows x64. We especially want real-world package, ODBC, authentication, browser, and workload feedback.
• Private Preview (invitation only): sqlsim and sqlpysim on macOS ARM64, and sqlsim Studio on Ubuntu x64, Windows ARM64, and macOS ARM64, while platform validation continues.
• Not currently supported: other Linux distributions and Linux ARM64. Reply here to tell us what you need so demand and prerequisites can be evaluated.

To ask for Private Preview access, open a Private Preview request. The form asks for the platform, SQL targets, authentication modes, and tools you can test, and access is granted to your GitHub account.

A request records interest; it does not guarantee access or establish a release date. Do not post email addresses, credentials, customer names, private server names, or other sensitive information. Maintainers will follow up through GitHub.

Propose a new workload

For a new simulation idea, describe:

• the problem or user scenario;
• the workload shape and target SQL products;
• the concurrency, correctness, latency, throughput, recall, resilience, or cost metrics that matter;
• the minimum useful output and reproducibility requirements; and
• why existing sqlsim or sqlpysim workload support is not enough.

Ideas such as independent TPC-like transactional or analytical patterns, vector search, RAG pipelines, multi-tenant SaaS load, connection storms, failover resilience, data movement, and event workloads are welcome. Benchmark-inspired proposals must use original schemas and scenarios and must not claim official benchmark compliance.

Pull requests during Public Preview

External pull requests are not accepted yet while interfaces, packaging, tests, and release processes stabilize. Please use Discussions for questions and early ideas, or Issues for agreed, trackable defects and feature requests. Maintainers may convert a mature discussion into an issue when its scope and acceptance criteria are clear.

Please follow the Microsoft Open Source Code of Conduct in every discussion.