Solana’s Alpenglow consensus upgrade is scheduled for Oct. 21, with activation now expected near 13:20 UTC after earlier estimates placed it at 14:50 UTC. The mainnet handoff will occur in epoch 1064, and developers are targeting roughly 150-millisecond finality under normal network conditions as Anza’s Votor mechanism replaces TowerBFT’s voting and finality logic.
Quicknode has advised operators to be ready from 13:00 UTC. During the transition, new transactions may briefly stop entering blocks for a few seconds—or longer if the handoff is delayed. Some submissions could temporarily return a “node unhealthy” preflight error, and Yellowstone data streams may disconnect once. The provider recommends using confirmed or finalized commitment for data that must not be reverted.
Validators must run Agave 4.3.2, Jito-Solana, or Harmonic Salsa before epoch 1064. Firedancer and Frankendancer are not supported during the migration. Operators also need at least 0.8 SOL above the rent-exempt minimum in vote accounts to cover the Validator Admission Ticket; an unpaid fee prevents voting or block production in epoch 1065.
The upgrade does not alter transaction execution, program behavior, or fee mechanics, and ordinary users will not need to change how they sign or approve transactions. A separate Rotor data-distribution upgrade remains unscheduled for later. Quicknode’s technical overview describes two voting paths: one can finalize a block in a single round when 80% of validators participate, while the other uses a second round at 60% participation. Before Alpenglow, applications often showed confirmations within two to three seconds, but full finality took about 13 seconds.
The consensus overhaul was approved through SIMD-0326 in September 2025 with 98.27% of votes in favor, representing 52% of stake. Alpenglow reached devnet on Sept. 26 after a testnet transition the previous day, while earlier in August its code was present in Agave 4.2 for testing without mainnet activation. Anza’s documentation warns that unstaked RPC nodes using the default delivery path may learn finality from certificates carried in later blocks, potentially adding up to one block interval compared with validators participating directly in consensus.
On the usage side, Solana reported 8.92 billion successful non-vote transactions in Q3 2026, up 104% from 4.38 billion in Q4 2025 and about 24% from Q2 2026. Independent data from Solana Compass recorded 8.78 billion completed transactions, a difference of roughly 140 million. The quarterly total implies an average of approximately 97 million transactions per day, or 1,122 per second, although the organic share of that activity remains unquantified.
SOL traded near $104.21 at the time of the market snapshot, up 3.3% over 24 hours but down 3.5% over seven days, with a market capitalization near $61.4 billion. DWF Labs research separately noted that Solana’s share of tracked tokenized-equity spot DEX volume fell to 31% in early October from 99.8% in January, but that data does not allocate the activity across networks or explain the transaction increase.
