Hyperledger · Rate Limits
Hyperledger Rate Limits
Hyperledger projects are self-hosted open-source software; there are no foundation-imposed request quotas. Operational throughput is bounded by node configuration, consensus algorithm, and underlying network/storage rather than a vendor-imposed rate-limit policy. Each framework exposes operator-configurable throttling controls (Fabric endorsement policies, Besu --rpc-http-max-active-connections, Firefly orchestrator backpressure). Managed offerings sold by ecosystem vendors apply their own service-specific limits.
5 Limits
Rate LimitingBlockchainOpen Source
Limits
Hyperledger Besu - JSON-RPC HTTP active connections node
80
Default --rpc-http-max-active-connections; configurable per-node.
Hyperledger Besu - JSON-RPC WebSocket active connections node
80
Default --rpc-ws-max-active-connections; configurable per-node.
Hyperledger Fabric - peer concurrent gRPC streams peer
operator-configured
Configured via core.yaml peer.keepalive and peer.maxRecvMsgSize parameters; no foundation-imposed ceiling.
Hyperledger Fabric - block size ordering-service
10485760
Default PreferredMaxBytes 512KB / AbsoluteMaxBytes 10MB at the orderer; configurable per channel.
Hyperledger Firefly - orchestration throughput node
implementation-dependent
Bounded by underlying blockchain protocol throughput and Firefly node sizing.
Policies
Self-Hosted Throttling
Operators set rate limits at the ingress layer (NGINX, Envoy, API gateway) since Hyperledger frameworks rarely ship with built-in per-client RPM throttling.
Consensus Backpressure
Throughput is gated by consensus mechanism (Raft, IBFT 2.0, QBFT, BFT). Excessive load causes block-time degradation rather than 429-style throttling.
Vendor Service Limits
Managed offerings (IBM Blockchain, AWS Managed Blockchain, Kaleido) impose their own rate limits, request quotas, and SLAs. Consult vendor docs for those specifics; the foundation itself does not.
Endorsement Policy
In Fabric, endorsement policies and concurrency are controlled at the chaincode and channel level rather than via API rate limits.