Solana Validator Glossary

Terms

Epoch

Solana groups blocks into 'epochs' — fixed windows of 432,000 slots that take about 2 days to complete on mainnet. Most validator metrics on this site are reported per epoch: an epoch closes, and its numbers become final.

Slot

A slot is the smallest unit of Solana's leader rotation — about 400ms. Each slot has exactly one validator scheduled as the leader; if that validator produces a block in the slot, it earns the block's fees.

Leader schedule

Before each epoch starts, Solana deterministically assigns leader slots to validators — biased by activated stake (more stake → more slots). The schedule is the source of truth for "which validator was supposed to produce block N".

Leaderboard window

The sample period used for a ranking. WhoEarns supports live_trend (running epoch plus latest final epoch), current_only, stable_trend, final_epoch, and decade_epoch. The same income / slot formula is applied inside each window.

Decade ranking

A leaderboard mode that uses the latest complete 10-epoch block. Validators need rows in all 10 epochs to qualify, and the Top 3 by income per leader slot receive gold, silver, and bronze badges.

Skip rate

The percentage of leader slots a validator failed to produce. 0% is perfect; consistently above ~5% usually means an unhealthy node or a misconfigured one. Skip rate is one of the cleanest signals of operator quality because it's stake-neutral.

Block fee

Total fees the leader receives for a block they produced. Decomposes into 'base fee' (a fixed 5,000 lamports per signature) and 'priority fee' (an optional extra users pay to outbid others for inclusion).

Priority fee

Optional extra users pay to outbid others for transaction inclusion in busy blocks. Since SIMD-96, 100% of priority fees go to the block leader (no burn).

MEV (Maximal Extractable Value)

Profit traders extract by re-ordering, inserting, or front-running transactions in a block. On Solana, most MEV flows through the Jito block engine — searchers bid for inclusion, the leader collects 'tips' on-chain.

Jito tips

On-chain SOL transfers to one of Jito's 8 tip accounts, deposited by traders to land bundles in this validator's blocks. This site derives MEV tips directly from produced block data.

Jito TipRouter

The on-chain program that distributes Jito MEV tips. It takes a protocol fee, then routes the rest to the validator and its delegators. WhoEarns reports the raw tips observed in produced blocks, not a delayed payout feed.

Lamport

The smallest unit of SOL. 1 SOL = 1,000,000,000 lamports. All raw on-chain amounts are lamports; this site shows both lamport (string for BigInt safety) and SOL (decimal) representations.

Vote account

The on-chain account stakers delegate to. A validator has exactly ONE vote account that lives across the lifetime of the validator — it's the canonical identifier and survives identity-key rotations.

Identity pubkey

The validator's hot key — signs blocks and votes. Operators may rotate this key periodically (common security practice) while keeping the same vote account.

Activated stake

The total SOL currently delegated to a validator and "active" (delegations have a one-epoch warmup). Drives leader-slot allocation: more activated stake = more leader slots = more income.

Commission

The percentage a validator takes from its delegators' INFLATION staking rewards before passing the rest on. Important: this site shows operator-side fee + tip income, which goes to the operator directly and is separate from the inflation rewards commission applies to. Commission is a delegator-facing cost; check it before delegating, but don't read this site's operator income as your yield.

APR / APY (operator)

Annualised return on activated stake from the operator side — what the validator's identity earns, NOT what a delegator receives. Calculated as (block fees + on-chain Jito tips) / activated stake × ~182 epochs/year. A delegator's own APR is a separate figure: it comes from inflation rewards minus the validator's commission, which this site does not track.

Reliability

A Node Tier sub-component: how dependably a validator produces the blocks it was scheduled to lead. Computed as 1 − the 95% Wilson UPPER bound on skip rate, so a small sample with zero observed skips can't claim a perfect score. A Wilson-upper skip rate above 20% hard-caps the tier at Kindling no matter how high the economic half is.

Economic percentile

A Node Tier sub-component: where a validator's median fee + tip income per leader slot ranks within the cohort of validators WhoEarns indexes — NOT against all of Solana. A percentile of 100% means top of the indexed cohort (which may be only a few dozen validators), not top of the whole cluster. Read the cohort size shown next to it.

CU subscore

A small (10%) part of the Node Tier economic half: the cohort percentile of a validator's average compute units packed into each produced block. It's a block-density lens — how much work per block — NOT a quality measure, and it never on its own forces an unrated tier. A validator that produced no blocks falls back to its economic percentile here.

Indexed average

The dashed peer line on the income and compute-unit charts: the MEAN per leader slot — income, or compute units per block — across the validators WhoEarns indexes that epoch (opt-outs excluded). NOT a whole-cluster figure. Note it's a different statistic from the Node Tier's Economic percentile, which ranks the MEDIAN: the chart uses a mean for a quick visual read; the tier uses a median because it's more robust for a score. A third 'same-client' line averages only the peers running the same validator client.

Missing a term? Send a PR on GitHub or ping 0base.vc.