QuanDeFi @QuanDeFi

Solana Data, explained by @waldruupi: “Solana Data is a product that came out of Solana to help align anybody who's confused by data and it serves two main audiences” "The first are the analysts and developers who constantly argue about what is actually going on. More importantly, the institutions trying to allocate capital." “This is something that came out of a need. It is not something that came out of us trying to think creatively."

Indexing years of transfers or replaying old market conditions gets expensive fast. With @Sam0xSol from @anza_xyz, we wrote a guide for the route that's free: 1. Backfill your own copy of Solana history with Jetstreamer, @anza_xyz's open-source replay engine 2. Feed it from Old Faithful, our public-good CAR archives of the full ledger from genesis 3. Stream everything into Superbank's ClickHouse schema with the ready-made plugin 4. Query history locally against your own storage, over JSON-RPC with gTFA, gRPC, or raw SQL Parallel decoding streams the data at over 4M TPS, with no rate limits and no per-request charges. Ready to try it out? Step-by-step walkthrough: blog.triton.one/self-hosting-s…

Cloudbreak delivers 99%+ faster getProgramAccounts on shared infrastructure. Instead of walking through over a billion accounts on every call, it jumps straight to the program's partition and reads a tailored index built for that exact query. It's live on every Triton endpoint, at the same price as a standard RPC call, and the whole pipeline is open source. Here's how we built it 👇

Superbank🦸🏦 is our open-source rebuild of Solana's entire history on ClickHouse, designed to make history performant, cost-efficient, and easy to use. - getSignatureStatuses: 38x faster - getSignaturesForAddress: 5x faster - getTransactionsForAddress: 1 call, 2x faster - Recent slots: <1ms (pass a slot hint to force it) We spent months building it around the queries you run most, so the looping, deduping, and custom workarounds stay out of your codebase 👇🧵

I'm super excited that the RPC Working Group has taken flight! The @solana read layer is finally IBRL-ing. We're expecting more open-source contributions from @JupiterExchange & @espicodes soon. Remember, kids -- it's open source, or it didn't happen!

Solana Developers @solana_devs

a month ago

Introducing the Solana RPC Working Group Its mission is to advance Solana's open-source RPC stack, making the read layer more accessible for all. Built in the open by @triton_one, @Helius, @JupiterExchange, @anza_xyz, and @SolanaFndn 🧵

16 43 267 493K 61

8/ Tx index metadata for ordered replay When datasources provide the block transaction index, v1 can now order activity by (slot, tx index) instead of slot alone. Which allows for better replay, more precise backfills and deterministic idempotency keys. github.com/sevenlabs-hq/c…

7/ Better CPI/event attribution correctness v1 improves CPI/event indexing by matching program logs to the right instruction, including precompile offsets that shift log positions. Now, more protocol activity lands under the correct tx and ix context. github.com/sevenlabs-hq/c…

5/ New slot-range & dedup filters Backfill windows and duplicate suppression no longer need custom datasource logic, processors, SQL, or DB constraints. Carbon now has reusable slot-range, tx-index-range, and dedup filters before storage. github.com/sevenlabs-hq/c…

4/ Cleaner pipeline observability Pipeline now owns routing, queueing, metrics, datasource identity, and shutdown behaviour–which allows for more operational visibility with less custom pipeline wiring. github.com/sevenlabs-hq/c…

3/ Exporter-based metrics Carbon moved metrics out of individual processor calls and into a central exporter model, so an indexer can expose runtime signals without hand-wiring metrics through every processing path. github.com/sevenlabs-hq/c…

2/ Smaller processor API Processors used to receive owned input data plus a metrics handle on every call; now they receive borrowed input references. This means smaller processor implementations and less framework-specific adapter code. github.com/sevenlabs-hq/c…

@SevenLabs_ just quietly released Carbon v1.0 Solana's most popular data indexing framework. Here are some of the biggest updates:

The full tutorial workflow can be found on the tutorial branch here: github.com/QuanDeFi/carbo… Tutorial Curriculum: github.com/QuanDeFi/carbo… Tutorial Recording Scripts: github.com/QuanDeFi/carbo…

And a computed branch diff document that shows the line by line and file by file difference to carbon v1: github.com/QuanDeFi/carbo…

The Carbon ClickHouse Sink has been fully tested and works end to end now with two examples: Jupiter-Swap settlements and USDC Token Account activity. The generated tutorial video shows the entire process in action.

Most of what teams want to build on @solana runs into the same wall: the read layer. RPC 2.0 takes it down. If you didn't make it to Accelerate, the full RPC 2.0 talk is now live on YouTube: youtu.be/eae22JihJ8Q. @brianlong covered: - Why the current monolith holds reads back, and how we fix it with the accounts and historical modules - What's behind it: independent storage, ingest, and query layers, purpose-built for one task - What opens up: you can run a custom index on shared infrastructure, easily query the full ledger since genesis, and prototype your app on a laptop

@triton_one @solana @brianlong good stuff and long overdue. when is the solana/rpc repo going public?

For years, custom indexes on Solana have meant paying for dedicated infrastructure. RPC 2.0 retires that trade-off and gives every builder adaptive indexes on the same standard JSON-RPC they already use.