Hackathon challenge: Scaling the DAO and BSQ

[HenrikJannsen]

DAO Performance Bottlenecks and Next Steps

The Bisq DAO was launched six years ago, and over time, the aggregated data (filtered blockchain data used for BSQ transactions) has become increasingly burdensome in terms of memory usage and performance for Bisq 1.

🔍 Analysis First

Before making architectural changes, we need a thorough analysis of the current pain points:

• Where exactly is performance lagging?
• Which parts of the DAO logic or data structures contribute most to memory bloat?

This analysis will inform the design of a more efficient and modular solution.

🏗️ Architectural Redesign

Once we’ve identified the issues, the next step is to design a cleaner, more scalable architecture. However, this must be done very carefully, as any changes must preserve DAO consensus—which is critical for the DAO's integrity.

🧩 Headless DAO Node

The goal is to refactor the DAO logic into a headless standalone node that exposes a gRPC or REST API. This would allow both:

• Bisq 1 to access DAO data via the new service, and
• Bisq 2 to consume it via the Oracle Node, without direct integration.

This approach would significantly reduce the memory and performance load on the client side and make the DAO logic easier to maintain and evolve.

📚 Getting Started

The DAO code currently lives in the Bisq 1 codebase. To become familiar with it:

• Start by reading the Bisq Wiki DAO documentation.
• Then explore the DAO-related classes and modules within the Bisq 1 source code.

Understanding the existing logic is essential before attempting any refactor or architectural redesign.

[skaunov]

Am I correct the linked page was moved to https://bisq.wiki/DAO_technical_overview?

[HenrikJannsen]

Yes. If you are interested to work on that, best to get in touch on matrix. My handle there is @boilingfrog:matrix.org
Our dev roadmap has changed a bit in the meantime and this task is quite large and would require quite some time from Bisq devs to assist and give background. We are atm fully focussed to get MuSig finished, thus everything else is deferred for post MuSig launch time.