tursodatabase/turso · error

--queries requires --benchmark search

Error message

--queries requires --benchmark search

What it means

The throughput benchmark is time-based and derives its workload from --seconds, not a query count. ensure! at perf/fts/src/main.rs:93 rejects --queries when --benchmark throughput is selected, since a fixed query count contradicts the run-until-duration model.

Solutions

1. Remove --queries when running --benchmark throughput
2. Keep --queries and switch back to --benchmark search (the default)
3. Make the wrapper script pass --queries only for search mode

Example fix

// before
fts-bench --benchmark throughput --queries 1000 --seconds 10
// after
fts-bench --benchmark throughput --seconds 10

if benchmark == Throughput && queries.is_some() { panic!("--queries only applies to the search benchmark"); }

When it happens

Trigger: Running `fts-bench --benchmark throughput --queries 1000`: ensure! fires because args.queries.is_some() in the Throughput branch. Happens whenever a search-style invocation is switched to throughput without dropping --queries.

Common situations: Reusing a search-benchmark command line for throughput runs; scripts parameterizing both --queries and --seconds and passing both unconditionally.

Understand the failure class

Background: "mutually exclusive" flag errors: what "can't supply both nx and xx", "--raw is not compatible with -i" and "cannot be used with" mean, and how to fix them — this error's family across 29 libraries.

Related errors

• first-query latency requires one query per connection
• need at least one query per connection
• seconds must be finite and positive
• --seconds requires --benchmark throughput
• throughput requires --state warm

AI-assisted analysis of tursodatabase/turso@8d4a589f8d (2026-09-20). Data as JSON: /api/errors/a265ff9bed991f80. Report an issue: GitHub.

Appendix: source

Thrown at perf/fts/src/main.rs:93

            let queries = args
                .queries
                .unwrap_or(if matches!(args.state, QueryState::First) {
                    args.connections
                } else {
                    10_000
                });
            ensure!(
                queries >= args.connections,
                "need at least one query per connection"
            );
            ensure!(
                !matches!(args.state, QueryState::First) || queries == args.connections,
                "first-query latency requires one query per connection"
            );
            (queries, 0.0)
        }
        Benchmark::Throughput => {
            ensure!(
                args.queries.is_none(),
                "--queries requires --benchmark search"
            );
            ensure!(
                matches!(args.state, QueryState::Warm),
                "throughput requires --state warm"
            );
            let seconds = args.seconds.unwrap_or(5.0);
            ensure!(
                seconds.is_finite() && seconds > 0.0,
                "seconds must be finite and positive"
            );
            (0, seconds)
        }
    };
    #[cfg(all(feature = "flamegraph", any(target_os = "linux", target_os = "macos")))]
    let profiled = args.flamegraph.is_some();
    #[cfg(not(all(feature = "flamegraph", any(target_os = "linux", target_os = "macos"))))]

View on GitHub (pinned to 8d4a589f8d)