[SOLO] Docker-based Bitcoin Full Node Mining Stack (Bitcoind + CKSolo + CKStats)

[MagicDude4Eva]

Just wanted to share a little project I’ve been working on over the past weeks. I’ve put together a full solo-mining stack built around Docker that runs well even on fairly modest hardware (in my case: a Synology DS1019+ with NVMe).

The goal: Run everything locally and hook it up to a NerdQaxe (or similar) for solo mining without relying on external services.

Included components:

• Bitcoin Core (bitcoind) – Full node
• CKPool Solo – with CKStats web UI
• Fulcrum – Electrum-compatible indexer (optional)
• Bitcoin RPC Explorer – simple web interface for Core
• InfluxDB + Grafana – real-time mining dashboard
• Mempool.space frontend – also working, but mostly just for fun

⚠️ Heads up: Fulcrum sync takes a long time, especially the first time. On my Synology DS1019+ with NVMe, it still took over a week. Stopping it midway will probably corrupt the index and you'll have to start over.

It’s all Dockerized, fairly easy to set up, and uses environment variables to wire everything together. The setup.sh script handles the permissions, clones the repo, sets up configs, etc. It’s opinionated, yes, but works pretty well out of the box.

🔗 GitHub: https://github.com/magicdude4eva/btc-fullnode-stack

If you try it out, let me know how it goes or if you run into issues. This is still experimental, but it’s been running stable here for over a week now.

[mikeywith]

without relying on external services.

You do realize that it's safer to rely on a stable external service than to run your own when it comes to mining? And by safer I am talking about your ability to actually win a block.

People often talk about bitcoin mining as if it's a race, which isn't entirely true. in reality, up to the point of when you find a valid hash, you are not racing against anyone, and your chances of hitting a block are independent from other miners hitting a block; however, once you do hit a block, the race starts. You need to make sure you propagate the block in a fraction of a second, mainly to other miners' nodes to let them know that you have already solved a block and that they need to start working on the next one. Other than that, it's possible that your block will be orphaned by another miner that has better connectivity to other nodes.

Add to it all the technical issues that your infrastructure can run into, like a small delay in downloading new blocks and clearning your mempool, and your getblocktemplate creates an invalid block. It takes a few components working together to ensure that your final constructed block is valid and propagated well and fast. everything must have 99.99% uptime, or else you are adding more risk for potentially no reward. You can't meaningfully have well-set-up mining infrastructure at home where your cat or kid may unplug the router power cable.

With that said, I do encourage everyone to test local mining for fun and educational purposes -- it's interesting to say the least, so good work you did there.

[MagicDude4Eva]

You do realize that it's safer to rely on a stable external service than to run your own when it comes to mining? And by safer I am talking about your ability to actually win a block.

Thanks for the feedback, you are absolutely right about the propagation and infrastructure risks. With the very-very-very slim chance of hitting a block in lottery mining, having it orphaned because the cat decided to nap in the warm air behind my rack and unplugging the fibre-link would be a true disaster. 😅

Your points are exactly the kind of things people should consider before going the self-hosted route. For me, it is part hobby, part infrastructure-geeking, and part dream of defying the odds.

That said, I have put some thought into stability: I run everything in a full-sized rack with a UPS, UniFi gear, PV + backup battery, fibre uplinks, and LTE failover. Running everything off NVMe with plenty of RAM has kept speeds and latency in a very comfortable range and I am genuinely happy with how stable it's been so far.

I do think that my Synology DS1019+ deserves an upgrade, but I do notice that the bitcoin-node publishes via ZMQ updates rather quick. I can not speak for the quality of work the miner produces and ckpool handles, but in the 2 days run-time the miner did 507M shares (3M rejection) and has an average hashrate of 4.3TH/s (it is a NerdQaxe++).