Skip to content

v0.15.0

Latest

Choose a tag to compare

@rtsaimonad rtsaimonad released this 30 Jun 16:53

⚠️ Do not upgrade your mainnet or testnet node until you receive an official announcement.
Upgrades are communicated via official channels. See upgrade instructions for details.

v0.15.0

Tags or hashes:

Highlights

  • MIP-12: Vote pace + block reward activation — updated vote pacing and block reward activation schedule
  • One-time DB metadata migration required (monad-mpt --upgrade) — bumps DB schema MONAD007 → MONAD008, initializing secondary timeline metadata for upcoming dual-timeline support
  • RPC batch concurrency limit — new server-side cap on concurrent batch requests to prevent resource exhaustion
  • eth_simulateV1 gated behind CLI flag — disabled by default; enable with --enable-eth-simulate-v1
  • Raptorcast starvation fix — networking message starvation in raptorcast resolved

Notable consensus/protocol changes (→ MINOR)

  • [Consensus] MIP-12: Update vote pace and block reward activation

Notable RPC/SDK changes

  • [RPC] Add batch concurrent limit — server-side cap on concurrent RPC batch requests
  • [RPC] Remove decompression support from RPC — compressed request bodies no longer accepted
  • [RPC] Gate eth_simulateV1 behind --enable-eth-simulate-v1 CLI flag (default: off) — returns "Method not supported" unless flag is set

Notable robustness changes

  • [Consensus] Fix networking message starvation in raptorcast
  • [Execution] Convert user-facing MONAD_ASSERT to MONAD_ASSERT_THROW — prevents panics on malformed RPC input
  • [Execution] Enforce staking syscall ordering — validation now rejects staking syscalls submitted out of order
  • [Execution] Fix event payload overflow check — rounds up size before alignment check to prevent assertion failure on large payloads
  • [Execution] Fix MONAD007→MONAD008 migration data corruption on nodes with ≥15.8 TB disk — prevents db_offsets overrun and root-offsets ring bound violation during monad-mpt --upgrade
  • [Execution] Fix monad-mpt --upgrade crash on legacy databases — fixes integer underflow causing Assertion 'r != MAP_FAILED' on nodes with databases created before num_cnv_chunks was added to the storage pool footer

Notable internal changes

Full Changelog: v0.14.7...v0.15.0