Unified data infrastructure
GitHub Let's talk

Performance guide

How to think about PLOMID performance — no invented numbers.

Version
Latest
v0.1.0 · latest 1 min read
On this page
  1. Levers you control

No universal benchmark numbers are claimed. Repository benches and perf probes are development measurements under their own conditions — cite conditions when quoting them.

Levers you control#

  1. Index the predicate: equality/range/order/time-window columns first; expression indexes for hot JSON paths.
  2. Shape the query: selective WHERE before joins; LIMIT early; prefer keyset over large OFFSET; EXPLAIN ANALYZE actuals over estimates.
  3. Batch writes: multi-row VALUES, INSERT…SELECT, short transactions; group commit already coalesces concurrent commits.
  4. Mind checkpoints: 64 MiB / 16 segments / 15 min defaults bound replay; raising thresholds defers checkpoint cost but lengthens recovery.
  5. Measure: ./target/debug/plomid benchmark run --iterations N --json writes <data_dir>/benchmarks/<run>.json (avg/p50/p95/p99/throughput for SELECT 1).

./target/debug/plomid metrics reports probe latency; server counters are unavailable in v0.1.0 (no metrics endpoint).

Was this page helpful?
Next

Continue

Next: SQL cookbook Practical recipes that teach how to think with PLOMID. Playground Run the documented statements against the sample dataset. Architecture The layers underneath the documented surface.