Rate Limits & Best Practices
Orbion’s public RPC endpoints are a shared resource provided for developers, testnet validators, and dApp builders. To ensure fair usage and stable performance across the ecosystem, requests are rate-limited and monitored for abusive behavior.
This page outlines current limits, usage guidelines, and recommended patterns for efficient integration.
📊 Rate Limit Policy
Requests per IP
20 requests per second (burstable)
Daily Cap
Soft cap at 250,000 req/day per IP
Payload Size
Max 1MB per request body
Concurrent Requests
5 per session recommended
WebSocket
Not supported yet
❗ Exceeding limits will result in temporary throttling (HTTP 429), progressive backoff, or IP cooldown.
⚠️ Throttling Behavior
When a client exceeds the permitted rate:
Requests may be delayed automatically (rate shaping)
Burst requests are deprioritized
IP cooldown is triggered on persistent abuse (up to 60 min)
Monitoring systems will blacklist aggressive bots or loops
✅ Best Practices for Developers
Follow these recommendations to avoid issues and improve your dApp’s RPC performance:
1. Avoid Repetitive Polling
Don't poll
eth_blockNumber
every 100ms — it wastes quota ⟶ Use block subscription (future) or check every 5–10s
2. Batch Requests When Possible
Combine multiple
eth_call
s into one batch usingeth_batch
3. Use Caching for Static Reads
Cache contract metadata,
eth_getCode
, token symbols, etc. ⟶ Reduce load and speed up UX
4. Back Off Gracefully
Implement exponential backoff when receiving HTTP 429 ⟶ Respect server load and retry later
5. Separate Frontend & Backend Traffic
Use server-side proxy for backend interactions if needed ⟶ Prevent client-side rate collisions
🧠 When to Run Your Own Node
If you are:
Indexing large sets of logs
Running dashboards or explorers
Building bots, aggregators, or bridge relayers
Reaching >500k RPC calls/day
Then it's highly recommended to operate a full or archive Orbion node. See Running a Node for instructions.
Orbion’s public RPC is designed to accelerate early development — not to replace dedicated infrastructure. Please use it respectfully and help maintain network health for everyone.
Last updated