Infrastructure
MegaETH solved the slow-chain problem. Next is the slow-infra problem. Oracles, indexers, apps—real-time chains demand innovation on traditional tools before users can feel real-time.
The Infrastructure Challenge
Traditional blockchain infrastructure assumes multi-second block times:
RPCs poll for receipts every few seconds
Indexers batch process blocks periodically
Oracles update prices on minute intervals
Apps show "pending" states for seconds
With 10ms blocks and instant execution, this infrastructure becomes the bottleneck. Users won't feel real-time performance if the tools can't keep up.
Realtime API
MegaETH's extension to Ethereum JSON-RPC optimized for sub-10ms latency.
Core Innovation: Query against mini-blocks, not EVM blocks
Transactions reflected within 10ms of arrival
State changes visible immediately
Full preconfirmation guarantees maintained
Enhanced Methods:
State queries (
eth_getBalance
,eth_call
, etc.):Return results up to most recent mini-block
Use
pending
orlatest
tag for real-time data100x faster state visibility vs waiting for EVM blocks
Transaction queries (
eth_getTransactionReceipt
):See transactions as soon as mini-block includes them
No more waiting for next EVM block
Receipt available in ~10ms
WebSocket subscriptions (
eth_subscribe
):Logs: Stream as transactions execute
State changes: Monitor account updates in real-time
Fragments: Subscribe to mini-block stream
Example timeline:
0ms: Transaction sent
10ms: Included in mini-block, visible in Realtime API
1000ms: Included in EVM block (traditional visibility)
realtime_sendRawTransaction
Revolutionary RPC method that bundles send + receipt in one call.
Problem it solves:
Traditional flow: Send tx → Poll for receipt → Multiple round trips
Polling wastes time between checks
Too frequent polling overloads infrastructure
Like checking mailbox every 10 minutes vs getting notification
How it works:
Send transaction to sequencer
Sequencer executes immediately (<10ms)
Returns receipt in same response
One RPC call for entire flow
Benefits:
Zero polling overhead
Minimal latency (network + 10ms execution)
Simple integration (drop-in replacement)
Better DevEx (no subscription management)
Implementation:
Paginated Reads
Robust API for processing massive chain data efficiently.
The data scale problem:
MegaETH testnet: 1000 TPS sustained
Generates 1 year of Ethereum data every 5 days
Traditional RPCs timeout on large queries
Apps forced to break queries into tiny chunks
Pagination solution:
Partial results: Return what's processed before limits hit
Resume pointers: Continue exactly where query stopped
No wasted work: Every computation counts
Optimal round trips: Minimize network overhead
Example use case: Query 1M blocks of logs:
Without pagination:
Query fails at 300k blocks
Retry with smaller range
Start from scratch each time
Unreliable, slow, complex
With pagination:
First call returns 300k blocks + pointer
Second call continues from block 300,001
Zero wasted computation
Predictable, fast, simple
Benefits for builders:
Indexers: Efficient backfilling after downtime
Analytics: Process entire chain history
Dashboards: Load large datasets reliably
Infrastructure: Handle data at scale
Infrastructure Roadmap
Current (Testnet):
✅ Realtime API with mini-block queries
✅ realtime_sendRawTransaction
✅ Paginated reads for logs
✅ WebSocket subscriptions
Coming Soon:
Extended pagination (traces, state)
Batch transaction submission
Priority transaction lanes
Historical state queries at mini-block level
Future Vision:
Sub-millisecond oracles
Streaming indexers
Real-time analytics
Instant cross-chain messaging
Builder Resources
Integration Guides:
SDK support coming for major libraries
Example implementations available
Best Practices:
Use
realtime_sendRawTransaction
for user transactionsSubscribe to mini-blocks for live data
Implement pagination for historical queries
Cache aggressively—data is immutable once confirmed
The infrastructure revolution is just beginning. As MegaETH pushes blockchain performance to new limits, expect innovative tools and services that reimagine what's possible in Web3.
Last updated