v0.7.1

• Fixed custom overlay msg sender

v0.7.0

Overview

QUIC-based public broadcast delivery, lower P2P and storage overhead, TVM optimizations and faster liteserver responses.

Changelog

Public broadcasts and P2P

• Added QUIC-based Plumtree delivery for public broadcasts, with FEC recovery for more reliable block propagation.
• Improved public broadcast routing to reduce duplicate traffic, bound memory usage, and isolate slow peers.
• Improved QUIC connectivity with cached routes, protected dialing, retry backoff, and cleanup of stale connections.
• Added persistent public-peer caching and a broader peer directory, allowing faster recovery after restart without keeping too many active connections.
• Hardened block broadcast processing: invalid and already-applied blocks are dropped early, while large payloads are decoded in bounded workers.
• Prioritized masterchain broadcast decoding so shard traffic does not delay the live masterchain head.
• Improved handling of out-of-order broadcasts by retrying valid payloads once their required predecessor state becomes available.

Synchronization and storage

• Avoided duplicate masterchain state preparation when several synchronization paths process the same block.
• Added early shard preparation so shard work can complete before its masterchain block is committed.
• Improved state-cell caching to keep large updates from blocking readers and other appliers for long periods.
• Improved archive catch-up by preparing the next archive window while the current one is being applied.
• Reworked archive import workers to prioritize active synchronization over background prefetch.
• Kept live block updates responsive when many recent blocks are still waiting to be saved.
• Reduced work and memory churn when publishing a block together with its state, especially on nodes with long block histories.
• Reduced temporary allocations while building shard-account state data, lowering memory pressure during large state operations.
• Improved persistent-state writes, checkpoint preparation, and Pebble batch handling to reduce unnecessary copying and buffer pressure.
• Improved Pebble compaction scheduling to avoid excessive retries and background work under storage pressure.

Liteserver, API, and operations

• Moved LiveView prewarming out of the block-apply path, keeping newly applied blocks available without waiting for expensive preparation.
• Added proactive liteserver warming for common requests at each new masterchain head, reducing cold-request latency.
• Improved historical block lookup through shard splits and merges, using live data before falling back to storage.
• Moved recent TPS calculation to a background tracker, reducing work during status and Prometheus collection.
• Enabled POST requests for all built-in HTTP API routes.
• Aligned rejected-external estimateFee responses with ton-http-api.
• Added mutex and block pprof profiles when --pprof-addr is enabled.
• Added node.Run(...) for custom node binaries using the standard CLI, configuration, and built-in extensions.

Upgrade notes

• The node now opens QUIC on UDP ADNL port + 1000; expose this port through firewall/NAT.
• gton_p2p_broadcasts_total gained a delivery label; update Prometheus rules and dashboards.
• Recent TPS metrics are published only after the live 10-second window is complete.

v0.6.0

Overview

Built-in ton-http-api compatible HTTP API, simplex block-finality broadcast support, TVM v15, optimized liteserver query execution, and major sync, P2P, cache, proof generation, and storage performance and reliability improvements.

Changelog

• Added an optional built-in HTTP API extension with /api/v2 GET, POST, and JSON-RPC support.
• Added tonNode.blockFinalityBroadcast support with simplex validator signature verification.
• Added independent caching and assembly of block candidates and finality broadcasts, regardless of arrival order.
• Added a bounded concurrent liteserver query dispatcher with fast paths, backlog shedding, cancellation handling, and larger per-connection response queues.
• Added liteserver.limits.max_waits_per_ip to limit parked waitMasterchainSeqno requests.
• Added full-block-identity response caching and shared in-flight computation for deterministic liteserver proofs, configuration, and shard data.
• Expanded shared block-proof construction and validation logic for reuse by liteserver, HTTP API, and broadcast processing.
• Added TVM v15 support and exposed the shared TVM instance to static extensions.
• Prioritized masterchain broadcasts in live-head sync
• Improved get-method execution, fee estimation, and external-message checking and broadcasting.
• Offloaded broadcast decompression and decoding to bounded workers with decoded-result caching.
• Added early rejection of already applied masterchain broadcasts and improved broadcast deduplication, payload reuse, and stale-message handling.
• Added active health probes and recovery for fixed custom overlays.
• Hardened peer endpoint identity, connection replacement, and stale peer cleanup.
• Reworked archive peer discovery and scouting with evidence-based admission, bounded DHT probing, per-demand health, retry backoff, and valuable peer reserves.
• Improved archive download sessions with shard-local peer selection, bounded retries, comparative hedging, subscription leases, cancellation, and idle pool cleanup.
• Improved archive import throughput with masterchain prefetching, zero-copy pack parsing, asynchronous artifact prewriting, and bounded buffering and backpressure.
• Added bounded streaming persistent-state downloads with advertised-size validation and disk-budget reservation.
• Parallelized state-cell imports and writes and optimized cell encoding, master-state caches, shard-state resolution, and transaction traversal.
• Hardened Pebble durability and recovery around artifact metadata, cell-generation rotation, persistent-state deletion, and archive pruning.
• Improved persistent-state retention for multiple records and masterchain groups while preserving compatibility with existing metadata.
• Hardened block identity validation across live caches, liteserver responses, state downloads, and stored artifacts.
• Improved non-final live-view publication with predecessor readiness checks, state-update validation reuse, and cheaper candidate-to-signed promotion.
• Optimized stored block replay by reusing verified artifacts and avoiding redundant file-hash checks and metadata rewrites.
• Updated the minimum Go version to 1.26 and bumped tonutils-go, RaptorQ, x/sys, CIRCL, and related dependencies.

v0.5.2

• No cell prewarm for archival live view publish

v0.5.1

• Support old global config without network id for DHT server

v0.5.0

Overview

Static node extensions and hooks support, liteserver refactored as a built-in extension, shared live-view store for liteserver and extensions, optional sync cutoff mode, checked external message flow, configurable local sendMessage fanout, storage/index migration, and multiple p2p, metrics, and storage reliability improvements.

Changelog

• Added static extension support through gton.RunNode, NodeOptions.Extension, and service/hooks.
• Added extension hooks for applied blocks, received blocks, and accepted external messages.
• Added extension capabilities for read-only live state access, checked external message sending, logging, and metrics.
• Added example_extension with apply logger and live transaction replay checker examples.
• Refactored liteserver into api/liteserver and wired the default liteserver as a static extension.
• Moved shared live liteserver storage/query logic into service/liveview for reuse by extensions.
• Added ton.sync_until to sync up to a UNIX block-time cutoff, then switch p2p offline while keeping local liteserver data available.
• Added liteserver.send_message_broadcast_fanout to tune local external-message broadcast fanout.
• Added TVM-based external message admission before local and overlay rebroadcast.
• Improved local sendMessage rebroadcasting with dedicated queues, configurable fanout and dedup, lag-aware fanout reduction.
• Added lazy state cell load metrics split by state window, decoded cache, and Pebble.
• Added incremental storage artifact metrics for archive packages and persistent state files.
• Improved p2p peer maintenance, peer warmup, pending peer promotion, and public-shard peer advertisement.
• Improved custom fixed overlay handling with lighter warmup and stricter advertised peer validation.
• Hardened archive package pruning to avoid deleting packages still pending sync.
• Hardened block metadata and block history indexes with full block-id hash validation.
• Added metadb v3 migration that rebuilds block history indexes from stored block metadata.
• Improved storage shutdown logging and close sequencing.
• Removed cmd/tvm-replay; replay checking is now provided as an extension example.
• Removed --from-zero and --liteserver-send-message-tvm-trace CLI flags.
• Changed archive-from-zero configuration to use ton.sync_before = -1.
• Updated tonutils-go.

v0.4.1

• Fixed block id copy in hot cache

v0.4.0

Overview

Archival node sync form zero state support, reduced block broadcasts apply latency (faster new blocks availability), faster archival blocks download, configurable liteserver rate limits, multiple performance optimizations.

Changelog

• Added archive node sync from zero support
• Improved archive downloads: peer pinning, cooldowns, retries, hedged downloads, and checkpoint backpressure.
• Improved broadcast flow: hot masterchain/shard caches, delayed compressed decode, and multi-block pending queues per chain.
• Added liteserver configurable rate limits: per-IP request limits, connection limits, keep-alive timeout.
• Improved live sync latency with priority masterchain broadcasts and faster catch-up from cached full blocks.
• Hardened state checkpoints: non-zero states now must be saved together with full block and proof artifacts.
• State snapshots no longer persist state-only checkpoints without block artifacts.
• Block history lookups now return only blocks with served-full artifacts.
• Hardened block metadata validation against forged served-full flags and direct next refs.
• Fixed checkpoint completion so older flushes do not drop newer applied states.
• Optimized liteserver sendMessage, account proof cache, block data loading, and non-final state cell lookup.
• Added configurable large-BOC read workers and persistent-state BOC batch size.
• Reduced Pebble cell-store allocations by reusing write batches.
• Improved liteserver errors parity with the C++ node.
• Metrics now expose applied masterchain progress separately from persisted state.
• Changed several defaults in config.
• Updated tonutils-go.

v0.3.3

• Fixed celldb migration state hole
• Optimized broadcast signatures verification
• Small refactor

v0.3.2

• Added --version flag for cmd/node, prints build GitCommit
• Improved live block sync responsiveness, pick up newly queued masterchain blocks faster.
• Reduced shard catch-up retry latency & broadcast event wake up
• Optimized archive catch-up memory usage: checkpointed applied states now keep metadata/root hash instead of retaining full cell graphs and parsed state.
• Improved archive checkpoint backpressure: pending checkpoint byte accounting now includes block/proof artifact payloads, not only state cell records.
• Reduced copying in archive state cell handling