feat(mergedmining): support dummy mining on coordinator

[jansegre]

Motivation

On the mainnet miners mostly use pools with custom integrations, but on the testnet the whole stack is in our control. For testing the newly activated (on the testnet) extended merkle_path len, we benefit from having the merged mining coordinator be able to do it seamlessly with some injected dummy merkle path.

Acceptance Criteria

• Add --x-dummy-merged-mining and --x-dummy-merkle-len to run_merged_mining subcommand, for running a coordinator with a dummy chain (instead of BTC), and with a dummy merkle path with the specified length;
• Adapt the merged mining coordinator to support mining without a BTC node and with dummy data instead.

Checklist

• If you are requesting a merge into master, confirm this code is production-ready and can be included in future releases as soon as it gets merged

[codecov]

Codecov Report

Attention: Patch coverage is 52.94118% with 16 lines in your changes missing coverage. Please review.

Project coverage is 84.99%. Comparing base (e61463e) to head (f1b24cd).

Files	Patch %	Lines
hathor/merged_mining/coordinator.py	54.54%	11 Missing and 4 partials ⚠️
hathor/client.py	0.00%	1 Missing ⚠️

Additional details and impacted files

@@            Coverage Diff             @@
##           master     #966      +/-   ##
==========================================
- Coverage   85.08%   84.99%   -0.10%     
==========================================
  Files         314      314              
  Lines       23919    23939      +20     
  Branches     3614     3621       +7     
==========================================
- Hits        20351    20346       -5     
- Misses       2863     2880      +17     
- Partials      705      713       +8     

☔ View full report in Codecov by Sentry.
📢 Have feedback on the report? Share it here.

[glevco]

I'm just wondering if this couldn't be simpler, for example, couldn't we just have a way to create a merge mined block and resolve it with the CPU miner? I mean, instead of having to start the coordinator, create a job, etc

[jansegre]

I'm just wondering if this couldn't be simpler, for example, couldn't we just have a way to create a merge mined block and resolve it with the CPU miner? I mean, instead of having to start the coordinator, create a job, etc

It definitely can be made simpler mining on CPU, however it's hard to make it in real-time to mine an actual block what will be accepted and used by the rest of the network. Which there wouldn't be anything wrong with a block that is voided for being old, but it'd be better if we have a more realistic test where blocks are accepted. Also doing it in the coordinator makes stratum clients mine blocks with the extended merkle path, which is good to confirm it works with no special changes.

[github-actions]

Bencher

Report	Wed, August 14, 2024 at 15:47:04 UTC
Project	hathor-core
Branch	feat/mm-coordinator-dummy
Testbed	ubuntu-22.04

Click to view all benchmark results

Benchmark	Latency	Latency Results nanoseconds (ns) | (Δ%)	Latency Lower Boundary nanoseconds (ns) | (%)	Latency Upper Boundary nanoseconds (ns) | (%)
sync-v2 (up to 20000 blocks)	✅ (view plot)	104,176,275,883.32 (-12.98%)	95,769,881,797.87 (91.93%)	143,654,822,696.81 (72.52%)

Bencher - Continuous Benchmarking
View Public Perf Page
Docs | Repo | Chat | Help