Running full nodes in production
Hey folks,
Looking to learn from mistakes/learnings of others, how are you running your full node of different crypto currencies in production? Inside of a Docker container? Some other way? Since the public ledger is often huge, have you worked out any means to share that across different instances/containers?
I imagine the only use-case is to scale JSON-RPC calls. If you use it in any other way, I would love to hear that :)
Trending on Indie Hackers
Depends on the use case, if you need historical data, then folks like casa send you historical data up to a date pre-loaded.
There are also hosted services like blockset, Google's OpenData set for a few base chains.
I am looking to run a full node myself, not really looking for an alternative solution. Also, for privacy reasons & not having to trust the 3rd party on the data, its always better to run the full node ourselves.
Hey did you figure something out, are you running parity nodes? This is a big issue for folks who would like to build decentralized from the ground up.
Not yet. Busy finishing my air-gapped wallet first. I am actually validating + building the entry to decentralized ecosystem from an individual perspective.
Yes. Checked out your progress. Nice work 👍.
Are you looking to implememt utrexxo?
You should probably look into crowdfunding the initial run
Thanks! I haven't looked into utrexxo yet. If its has real metric, I will eventually build it.
I would love for a model to fund my work so that I can open source everything that I build but I can't pull myself away from real work to figure out crowdfunding just yet. Not to mention if it will even work. Perhaps in next year when I would have some pillars to show for my dedication & vision.
Assume we're talking about PoS systems here, but running a full node is the easy part, validators are more difficult.
Something like docker yes with auto restarts and a mounted volume is what you want.
There are chains out there that allow you to run nodes in light mode where you just sync the headers of each block in an epoch, that allows you to run nodes on resource constrained devices while still accepting RPCs
More like enabling POS systems to use my API. By validator do you mean running one's own software layer on top to be able to query any data they want, like blockchain explorers do?
I am aware of SPV as well, but I am really looking to learn what it takes to scale a lot of JSON-RPC calls traffic. No idea how light weight they are or if scaling them is a challenge.
Can multiple instances share the same downloaded public ledger? Same problem as multiple write attempts on a single file. Or scaling vertically is the only option? Wonder how people are doing it.