PPoS / Consensus / DA §
Research §
- A document was prepared to analyze the data sizes when sending data chunk by chunk versus sending the whole column in preparation for the DA Sampling research. The document will be updated as the sampling research is examined in more detail.
- Sampling document has been updated but still WIP.
- Continued writing the report that compares the tx fee mechanisms and added some adaptations to the code.
Development §
Blendnet §
Research §
- Blendnet experiments - latency & bandwidth: wrote the draft report but performing additional detailed experiments to clearly understand the impact of parameters on latency. To share the intermediate results, 2-hop blending can be achieved with max 10s latency and max 150Kbps bandwidth (excluding broadcasting) when the network size is 1000. However, further analysis is required.
- Blendnet experiments - connection monitoring: wrote the report. Based on the analysis so far, it has been observed that a very small number of connections remain idle for a certain period, even when all nodes are stable. According to the current algorithm, such connections can easily be deemed ‘unhealthy’. Currently closely investigating the reasons for this, and if necessary, considering refining the unhealthy peer detection algorithm for greater precision.
- Integrated the connection monitor to the simulation: Developed in a branch but not opened a PR yet because some refactoring in the
nomos-node must be done first.
- The design of the Economics Motivation Module was finished. However, the document is still in WIP as it describes competing two approaches for emission quota registration and it is still needed to review constructions.
- The difference between registration approaches is (initially) discussed as the impact on the cost of the registration of the edge node.
- Updated the connection monitoring description as it was found during simulations that we had the same parameter for monitoring upper and lower bound on the expected emission. Now we have separated parameters which allows us to distinguish between malicious and unhealthy connections.
- Analysis of communication (and anonymity) failure in the Blendnet: considered broadcasting on linear trees where a message is sent through K paths of length L. For linear trees we have derived analytically equations for the probability of broadcasting and anonymity failures. Wrote a piece of simulation code, which assumes that M number of messages was sent, and verified the above analytic result by simulation. For the broadcasting on a branching tree, wrote simulation code and verified analytic results for the probability of broadcasting and anonymity failures by simulation. The simulation suggests that for finite number M deviations from analytic results, which assume infinite M, are observed. Thus analysis of these deviations has to be carried out. The summary of the above work (in progress) is in this document while the detailed analysis can be seen here.
Development §
Coordination Layer §
Research §
Development §
Zone and Apps §
Research §
Development §
Development §
- Currently in review: VAC/QA testing planning proposal and current status.
- PR #931 (Node) - tracing bandwidth metrics: DA.
- PR #933 (Node) - tracing bandwidth metrics: Blendnet.
- PR #934 (Node) - tracing bandwidth metrics: Consensus.
- SurrealDB proof of concept: cleaned up notes and added postponement issue.
- Added protobuf issue.
- Ran benchmarks - encoder.
- Ran benchmarks - verifier.
- Ran benchmarks - reconstruction.
- PR #942 (Node) - added
pre-commit hook configuration.
- WIP branch - split tracing and other updates from metrics: split the functionality that is working and configured correctly on the testnet host into a separate PR.
- WIP branch - investigating unreliable metrics on the testnet host, not all metrics are available on the host as in local development environment. There are quite a few moving parts, a bit more time is needed until its usable by the team.
Miscellaneous §
- zkVM survey published.
- NomosDA blog post.
- NomosDA video script. Waiting on reviews to be finalized.
- 2024 recap blog. (WIP)