Solana's Alpenglow Switch Could Reset Transaction Count Baselines

Solana's scheduled Alpenglow upgrade could lower reported transaction counts without a matching decline in user activity, requiring a new baseline for dashboards that include validator votes in their totals.

The planned mainnet upgrade is set to begin at epoch 1064, expected around 11 a.m. ET (15:00 UTC) on Oct. 21. That marks the start of the feature gate, not the consensus switch. Alpenglow consensus is expected to take over about 5,000 slots later, roughly 18 minutes after the epoch begins.

Under the current TowerBFT system, validators record their votes as transactions in blocks. Alpenglow's Votor protocol instead has validators exchange votes directly and combine them into certificates. If the upgrade proceeds, those votes will no longer appear among the transactions counted from blocks.

That change matters for transaction totals and transactions-per-second figures. A dashboard that counts validator votes alongside user transactions could show a sharp drop after the switch even if users continue sending transactions at the same rate. Comparisons across the upgrade will need to account for the change in what a block contains.

Tools that track validator participation through block transactions will also need to use a different record. Under Alpenglow, that information moves to certificates in the block footer. The change affects how participation is observed, as well as how overall transaction activity is counted.

Transaction execution and transaction formats are expected to remain the same. User transaction fee mechanics are also unchanged. The accounting shift therefore does not, by itself, indicate a change in how users submit transactions or what they pay to do so.

Alpenglow targets finality of roughly 150 milliseconds, compared with about 12.8 seconds under TowerBFT. Finality is the point at which a transaction is treated as irreversible. The faster figure is a target; actual mainnet finality and a comparable baseline for transaction counts can be assessed after the scheduled consensus switch.