Cardano Leios Benchmark Reach Stake Pool Operators

Operators fit test transaction validation without development toolchain, as the team dey find disk measurements from wider range of machines.

By SongMarketCap

Cardano News - Cardano Leios Benchmark Reach Stake Pool Operators

Cardano stake pool operators don receive standalone transaction validation benchmark, na wetin the Performance & Tracing team report on September 30, 2026. The executable go support hardware testing as developers dey prepare Ouroboros Leios, one proposed upgrade to Cardano transaction processing capacity.

Wetín the Cardano Leios Benchmark Dey Measure

Leios dey introduce extra blocks wey dey reference transactions alongside the existing Praos chain. One committee go verify these blocks before their transactions enter the ledger, so e go expand the amount of transaction data wey the network fit process.

The benchmark dey examine the ledger component, including transaction validation and state updates. Transaction and block propagation across live network no dey inside im scope.

Operators fit run the delivered executable without Nix or Haskell development environment. The team dey find measurements from different hardware configurations, especially those wey dey use the on-disk implementation of LedgerDB.

Why Disk Performance Matter for Leios

LedgerDB dey manage the state wey needed to process transactions, including unspent transaction outputs, or UTxOs. As that data dey stored for disk, to retrieve am go add storage access join the work wey the processor dey do.

Because of that, machine performance fit depend on im SSD, storage connection, operating system kernel, file system, and mount settings.

The beacon benchmarking tool get option to bypass the operating system page cache, wey go force the test to access storage directly. This one dey allow developers examine disk costs wey RAM fit hide if data dey stay for cache.

Im measurements include total elapsed time, time wey e take read UTxO tables, peak memory use, and disk I/O activity. Together, these metrics dey describe both processing time and the resources wey run don use.

Hardware Reports and the Public Glue Build

Public development work on glue dey document wrapper wey dey download one chain fragment, run four benchmark configurations, and assemble ZIP wey contain results, machine information, and analyzer build details. Linux builds cover x86_64 and ARM64. The workflow dey download approximately 800 MiB and require about 2.5 GiB of free disk space.

The public nightly na hardware-testing prerelease wey dem build from unreviewed draft code. Im release notes dey distinguish am from the SPO deliverable. The September 30 update no identify the version wey dem send give operators.

Before Leios release candidate, the team still dey test automation for performance cluster. The report talk say these runs dey prepare the infrastructure and no go produce representative Leios performance figures. The September 30 update no publish operator measurements or minimum hardware requirements for running Leios.