MeiliSearch/benchmarks/benches/indexing.rs

555 lines
22 KiB
Rust
Raw Normal View History

2021-07-07 11:42:14 +02:00
mod datasets_paths;
mod utils;
2021-07-07 11:42:14 +02:00
use std::fs::{create_dir_all, remove_dir_all};
2021-07-07 11:42:14 +02:00
use std::path::Path;
use criterion::{criterion_group, criterion_main, Criterion};
use heed::EnvOpenOptions;
use milli::update::{IndexDocuments, IndexDocumentsConfig, IndexerConfig, Settings};
2021-07-07 11:42:14 +02:00
use milli::Index;
#[cfg(target_os = "linux")]
#[global_allocator]
static ALLOC: jemallocator::Jemalloc = jemallocator::Jemalloc;
fn setup_dir(path: impl AsRef<Path>) {
match remove_dir_all(path.as_ref()) {
Ok(_) => (),
Err(e) if e.kind() == std::io::ErrorKind::NotFound => (),
Err(e) => panic!("{}", e),
}
create_dir_all(path).unwrap();
}
fn setup_index() -> Index {
let path = "benches.mmdb";
setup_dir(&path);
let mut options = EnvOpenOptions::new();
options.map_size(100 * 1024 * 1024 * 1024); // 100 GB
options.max_readers(10);
Index::new(options, path).unwrap()
}
fn indexing_songs_default(c: &mut Criterion) {
let mut group = c.benchmark_group("indexing");
group.sample_size(10);
group.bench_function("Indexing songs with default settings", |b| {
b.iter_with_setup(
move || {
2021-08-19 15:02:43 +02:00
let index = setup_index();
let config = IndexerConfig::default();
2021-08-19 15:02:43 +02:00
let mut wtxn = index.write_txn().unwrap();
let mut builder = Settings::new(&mut wtxn, &index, &config);
2021-08-19 15:02:43 +02:00
builder.set_primary_key("id".to_owned());
let displayed_fields =
["title", "album", "artist", "genre", "country", "released", "duration"]
.iter()
.map(|s| s.to_string())
.collect();
builder.set_displayed_fields(displayed_fields);
let searchable_fields =
["title", "album", "artist"].iter().map(|s| s.to_string()).collect();
builder.set_searchable_fields(searchable_fields);
let faceted_fields =
["released-timestamp", "duration-float", "genre", "country", "artist"]
.iter()
.map(|s| s.to_string())
.collect();
builder.set_filterable_fields(faceted_fields);
2021-11-03 13:12:01 +01:00
builder.execute(|_| ()).unwrap();
2021-07-07 11:42:14 +02:00
wtxn.commit().unwrap();
2021-08-19 15:02:43 +02:00
index
2021-07-07 11:42:14 +02:00
},
2021-08-19 15:02:43 +02:00
move |index| {
let config = IndexerConfig::default();
let indexing_config = IndexDocumentsConfig::default();
2021-08-19 15:02:43 +02:00
let mut wtxn = index.write_txn().unwrap();
let mut builder =
IndexDocuments::new(&mut wtxn, &index, &config, indexing_config, |_| ());
let documents = utils::documents_from(datasets_paths::SMOL_SONGS, "csv");
builder.add_documents(documents).unwrap();
builder.execute().unwrap();
2021-07-07 11:42:14 +02:00
wtxn.commit().unwrap();
2021-08-19 15:02:43 +02:00
index.prepare_for_closing().wait();
2021-07-07 11:42:14 +02:00
},
)
});
}
fn indexing_songs_in_three_batches_default(c: &mut Criterion) {
let mut group = c.benchmark_group("indexing");
group.sample_size(10);
group.bench_function("Indexing songs in three batches with default settings", |b| {
b.iter_with_setup(
move || {
let index = setup_index();
let config = IndexerConfig::default();
let mut wtxn = index.write_txn().unwrap();
let mut builder = Settings::new(&mut wtxn, &index, &config);
builder.set_primary_key("id".to_owned());
let displayed_fields =
["title", "album", "artist", "genre", "country", "released", "duration"]
.iter()
.map(|s| s.to_string())
.collect();
builder.set_displayed_fields(displayed_fields);
let searchable_fields =
["title", "album", "artist"].iter().map(|s| s.to_string()).collect();
builder.set_searchable_fields(searchable_fields);
let faceted_fields =
["released-timestamp", "duration-float", "genre", "country", "artist"]
.iter()
.map(|s| s.to_string())
.collect();
builder.set_filterable_fields(faceted_fields);
builder.execute(|_| ()).unwrap();
// We index only one half of the dataset in the setup part
2022-02-22 17:39:24 +01:00
// as we don't care about the time it takes.
let config = IndexerConfig::default();
let indexing_config = IndexDocumentsConfig::default();
let mut builder =
IndexDocuments::new(&mut wtxn, &index, &config, indexing_config, |_| ());
let documents = utils::documents_from(datasets_paths::SMOL_SONGS_1_2, "csv");
builder.add_documents(documents).unwrap();
builder.execute().unwrap();
wtxn.commit().unwrap();
index
},
move |index| {
let config = IndexerConfig::default();
let indexing_config = IndexDocumentsConfig::default();
let mut wtxn = index.write_txn().unwrap();
let mut builder =
IndexDocuments::new(&mut wtxn, &index, &config, indexing_config, |_| ());
let documents = utils::documents_from(datasets_paths::SMOL_SONGS_3_4, "csv");
builder.add_documents(documents).unwrap();
builder.execute().unwrap();
let indexing_config = IndexDocumentsConfig::default();
let mut builder =
IndexDocuments::new(&mut wtxn, &index, &config, indexing_config, |_| ());
let documents = utils::documents_from(datasets_paths::SMOL_SONGS_4_4, "csv");
builder.add_documents(documents).unwrap();
builder.execute().unwrap();
wtxn.commit().unwrap();
index.prepare_for_closing().wait();
},
)
});
}
2021-07-07 11:42:14 +02:00
fn indexing_songs_without_faceted_numbers(c: &mut Criterion) {
let mut group = c.benchmark_group("indexing");
group.sample_size(10);
group.bench_function("Indexing songs without faceted numbers", |b| {
b.iter_with_setup(
move || {
2021-08-19 15:02:43 +02:00
let index = setup_index();
let config = IndexerConfig::default();
2021-08-19 15:02:43 +02:00
let mut wtxn = index.write_txn().unwrap();
let mut builder = Settings::new(&mut wtxn, &index, &config);
2021-08-19 15:02:43 +02:00
builder.set_primary_key("id".to_owned());
let displayed_fields =
["title", "album", "artist", "genre", "country", "released", "duration"]
.iter()
.map(|s| s.to_string())
.collect();
builder.set_displayed_fields(displayed_fields);
let searchable_fields =
["title", "album", "artist"].iter().map(|s| s.to_string()).collect();
builder.set_searchable_fields(searchable_fields);
let faceted_fields =
["genre", "country", "artist"].iter().map(|s| s.to_string()).collect();
builder.set_filterable_fields(faceted_fields);
2021-11-03 13:12:01 +01:00
builder.execute(|_| ()).unwrap();
2021-07-07 11:42:14 +02:00
wtxn.commit().unwrap();
2021-08-19 15:02:43 +02:00
index
2021-07-07 11:42:14 +02:00
},
2021-08-19 15:02:43 +02:00
move |index| {
let config = IndexerConfig::default();
let indexing_config = IndexDocumentsConfig::default();
2021-08-19 15:02:43 +02:00
let mut wtxn = index.write_txn().unwrap();
let mut builder =
IndexDocuments::new(&mut wtxn, &index, &config, indexing_config, |_| ());
let documents = utils::documents_from(datasets_paths::SMOL_SONGS, "csv");
builder.add_documents(documents).unwrap();
builder.execute().unwrap();
2021-07-07 11:42:14 +02:00
wtxn.commit().unwrap();
2021-08-19 15:02:43 +02:00
index.prepare_for_closing().wait();
2021-07-07 11:42:14 +02:00
},
)
});
}
fn indexing_songs_without_faceted_fields(c: &mut Criterion) {
let mut group = c.benchmark_group("indexing");
group.sample_size(10);
group.bench_function("Indexing songs without any facets", |b| {
b.iter_with_setup(
move || {
2021-08-19 15:02:43 +02:00
let index = setup_index();
let config = IndexerConfig::default();
2021-08-19 15:02:43 +02:00
let mut wtxn = index.write_txn().unwrap();
let mut builder = Settings::new(&mut wtxn, &index, &config);
2021-08-19 15:02:43 +02:00
builder.set_primary_key("id".to_owned());
let displayed_fields =
["title", "album", "artist", "genre", "country", "released", "duration"]
.iter()
.map(|s| s.to_string())
.collect();
builder.set_displayed_fields(displayed_fields);
let searchable_fields =
["title", "album", "artist"].iter().map(|s| s.to_string()).collect();
builder.set_searchable_fields(searchable_fields);
2021-11-03 13:12:01 +01:00
builder.execute(|_| ()).unwrap();
2021-07-07 11:42:14 +02:00
wtxn.commit().unwrap();
2021-08-19 15:02:43 +02:00
index
2021-07-07 11:42:14 +02:00
},
2021-08-19 15:02:43 +02:00
move |index| {
let config = IndexerConfig::default();
let indexing_config = IndexDocumentsConfig::default();
2021-08-19 15:02:43 +02:00
let mut wtxn = index.write_txn().unwrap();
let mut builder =
IndexDocuments::new(&mut wtxn, &index, &config, indexing_config, |_| ());
let documents = utils::documents_from(datasets_paths::SMOL_SONGS, "csv");
builder.add_documents(documents).unwrap();
builder.execute().unwrap();
2021-07-07 11:42:14 +02:00
wtxn.commit().unwrap();
2021-08-19 15:02:43 +02:00
index.prepare_for_closing().wait();
2021-07-07 11:42:14 +02:00
},
)
});
}
fn indexing_wiki(c: &mut Criterion) {
let mut group = c.benchmark_group("indexing");
group.sample_size(10);
group.bench_function("Indexing wiki", |b| {
b.iter_with_setup(
move || {
2021-08-19 15:02:43 +02:00
let index = setup_index();
let config = IndexerConfig::default();
2021-08-19 15:02:43 +02:00
let mut wtxn = index.write_txn().unwrap();
let mut builder = Settings::new(&mut wtxn, &index, &config);
2021-08-19 15:02:43 +02:00
builder.set_primary_key("id".to_owned());
let displayed_fields =
["title", "body", "url"].iter().map(|s| s.to_string()).collect();
builder.set_displayed_fields(displayed_fields);
let searchable_fields = ["title", "body"].iter().map(|s| s.to_string()).collect();
builder.set_searchable_fields(searchable_fields);
// there is NO faceted fields at all
2021-11-03 13:12:01 +01:00
builder.execute(|_| ()).unwrap();
2021-07-07 11:42:14 +02:00
wtxn.commit().unwrap();
2021-08-19 15:02:43 +02:00
index
2021-07-07 11:42:14 +02:00
},
2021-08-19 15:02:43 +02:00
move |index| {
let config = IndexerConfig::default();
let indexing_config =
IndexDocumentsConfig { autogenerate_docids: true, ..Default::default() };
2021-08-19 15:02:43 +02:00
let mut wtxn = index.write_txn().unwrap();
let mut builder =
IndexDocuments::new(&mut wtxn, &index, &config, indexing_config, |_| ());
let documents = utils::documents_from(datasets_paths::SMOL_WIKI_ARTICLES, "csv");
builder.add_documents(documents).unwrap();
builder.execute().unwrap();
2021-07-07 11:42:14 +02:00
wtxn.commit().unwrap();
2021-08-19 15:02:43 +02:00
index.prepare_for_closing().wait();
2021-07-07 11:42:14 +02:00
},
)
});
}
fn indexing_wiki_in_three_batches(c: &mut Criterion) {
let mut group = c.benchmark_group("indexing");
group.sample_size(10);
group.bench_function("Indexing wiki in three batches", |b| {
b.iter_with_setup(
move || {
let index = setup_index();
let config = IndexerConfig::default();
let mut wtxn = index.write_txn().unwrap();
let mut builder = Settings::new(&mut wtxn, &index, &config);
builder.set_primary_key("id".to_owned());
let displayed_fields =
["title", "body", "url"].iter().map(|s| s.to_string()).collect();
builder.set_displayed_fields(displayed_fields);
let searchable_fields = ["title", "body"].iter().map(|s| s.to_string()).collect();
builder.set_searchable_fields(searchable_fields);
// there is NO faceted fields at all
builder.execute(|_| ()).unwrap();
// We index only one half of the dataset in the setup part
2022-02-22 17:39:24 +01:00
// as we don't care about the time it takes.
let config = IndexerConfig::default();
let indexing_config =
IndexDocumentsConfig { autogenerate_docids: true, ..Default::default() };
let mut builder =
IndexDocuments::new(&mut wtxn, &index, &config, indexing_config, |_| ());
let documents =
utils::documents_from(datasets_paths::SMOL_WIKI_ARTICLES_1_2, "csv");
builder.add_documents(documents).unwrap();
builder.execute().unwrap();
wtxn.commit().unwrap();
index
},
move |index| {
let config = IndexerConfig::default();
let indexing_config =
IndexDocumentsConfig { autogenerate_docids: true, ..Default::default() };
let mut wtxn = index.write_txn().unwrap();
let mut builder =
IndexDocuments::new(&mut wtxn, &index, &config, indexing_config, |_| ());
let documents =
utils::documents_from(datasets_paths::SMOL_WIKI_ARTICLES_3_4, "csv");
builder.add_documents(documents).unwrap();
builder.execute().unwrap();
let indexing_config =
IndexDocumentsConfig { autogenerate_docids: true, ..Default::default() };
let mut builder =
IndexDocuments::new(&mut wtxn, &index, &config, indexing_config, |_| ());
let documents =
utils::documents_from(datasets_paths::SMOL_WIKI_ARTICLES_4_4, "csv");
builder.add_documents(documents).unwrap();
builder.execute().unwrap();
wtxn.commit().unwrap();
index.prepare_for_closing().wait();
},
)
});
}
2021-07-07 11:42:14 +02:00
fn indexing_movies_default(c: &mut Criterion) {
let mut group = c.benchmark_group("indexing");
group.sample_size(10);
group.bench_function("Indexing movies with default settings", |b| {
b.iter_with_setup(
move || {
2021-08-19 15:02:43 +02:00
let index = setup_index();
let config = IndexerConfig::default();
2021-08-19 15:02:43 +02:00
let mut wtxn = index.write_txn().unwrap();
let mut builder = Settings::new(&mut wtxn, &index, &config);
2021-08-19 15:02:43 +02:00
builder.set_primary_key("id".to_owned());
let displayed_fields = ["title", "poster", "overview", "release_date", "genres"]
.iter()
.map(|s| s.to_string())
.collect();
builder.set_displayed_fields(displayed_fields);
let searchable_fields =
["title", "overview"].iter().map(|s| s.to_string()).collect();
builder.set_searchable_fields(searchable_fields);
let faceted_fields =
["released_date", "genres"].iter().map(|s| s.to_string()).collect();
builder.set_filterable_fields(faceted_fields);
2021-11-03 13:12:01 +01:00
builder.execute(|_| ()).unwrap();
2021-07-07 11:42:14 +02:00
wtxn.commit().unwrap();
2021-08-19 15:02:43 +02:00
index
2021-07-07 11:42:14 +02:00
},
2021-08-19 15:02:43 +02:00
move |index| {
let config = IndexerConfig::default();
let indexing_config = IndexDocumentsConfig::default();
2021-08-19 15:02:43 +02:00
let mut wtxn = index.write_txn().unwrap();
let mut builder =
IndexDocuments::new(&mut wtxn, &index, &config, indexing_config, |_| ());
2021-07-07 11:42:14 +02:00
let documents = utils::documents_from(datasets_paths::MOVIES, "json");
builder.add_documents(documents).unwrap();
builder.execute().unwrap();
2021-07-07 11:42:14 +02:00
wtxn.commit().unwrap();
2021-08-19 15:02:43 +02:00
index.prepare_for_closing().wait();
2021-07-07 11:42:14 +02:00
},
)
});
}
fn indexing_movies_in_three_batches(c: &mut Criterion) {
let mut group = c.benchmark_group("indexing");
group.sample_size(10);
group.bench_function("Indexing movies in three batches", |b| {
b.iter_with_setup(
move || {
let index = setup_index();
let config = IndexerConfig::default();
let mut wtxn = index.write_txn().unwrap();
let mut builder = Settings::new(&mut wtxn, &index, &config);
builder.set_primary_key("id".to_owned());
let displayed_fields = ["title", "poster", "overview", "release_date", "genres"]
.iter()
.map(|s| s.to_string())
.collect();
builder.set_displayed_fields(displayed_fields);
let searchable_fields =
["title", "overview"].iter().map(|s| s.to_string()).collect();
builder.set_searchable_fields(searchable_fields);
let faceted_fields =
["released_date", "genres"].iter().map(|s| s.to_string()).collect();
builder.set_filterable_fields(faceted_fields);
builder.execute(|_| ()).unwrap();
// We index only one half of the dataset in the setup part
2022-02-22 17:39:24 +01:00
// as we don't care about the time it takes.
let config = IndexerConfig::default();
let indexing_config = IndexDocumentsConfig::default();
let mut builder =
IndexDocuments::new(&mut wtxn, &index, &config, indexing_config, |_| ());
let documents = utils::documents_from(datasets_paths::MOVIES_1_2, "json");
builder.add_documents(documents).unwrap();
builder.execute().unwrap();
wtxn.commit().unwrap();
index
},
move |index| {
let config = IndexerConfig::default();
let indexing_config = IndexDocumentsConfig::default();
let mut wtxn = index.write_txn().unwrap();
let mut builder =
IndexDocuments::new(&mut wtxn, &index, &config, indexing_config, |_| ());
let documents = utils::documents_from(datasets_paths::MOVIES_3_4, "json");
builder.add_documents(documents).unwrap();
builder.execute().unwrap();
let indexing_config = IndexDocumentsConfig::default();
let mut builder =
IndexDocuments::new(&mut wtxn, &index, &config, indexing_config, |_| ());
let documents = utils::documents_from(datasets_paths::MOVIES_4_4, "json");
builder.add_documents(documents).unwrap();
builder.execute().unwrap();
wtxn.commit().unwrap();
index.prepare_for_closing().wait();
},
)
});
}
2021-09-13 18:08:28 +02:00
fn indexing_geo(c: &mut Criterion) {
let mut group = c.benchmark_group("indexing");
group.sample_size(10);
group.bench_function("Indexing geo_point", |b| {
b.iter_with_setup(
move || {
let index = setup_index();
let config = IndexerConfig::default();
2021-09-13 18:08:28 +02:00
let mut wtxn = index.write_txn().unwrap();
let mut builder = Settings::new(&mut wtxn, &index, &config);
2021-09-13 18:08:28 +02:00
builder.set_primary_key("geonameid".to_owned());
let displayed_fields =
["geonameid", "name", "asciiname", "alternatenames", "_geo", "population"]
.iter()
.map(|s| s.to_string())
.collect();
builder.set_displayed_fields(displayed_fields);
let searchable_fields =
["name", "alternatenames", "elevation"].iter().map(|s| s.to_string()).collect();
builder.set_searchable_fields(searchable_fields);
let filterable_fields =
["_geo", "population", "elevation"].iter().map(|s| s.to_string()).collect();
builder.set_filterable_fields(filterable_fields);
let sortable_fields =
["_geo", "population", "elevation"].iter().map(|s| s.to_string()).collect();
builder.set_sortable_fields(sortable_fields);
2021-11-03 13:12:01 +01:00
builder.execute(|_| ()).unwrap();
2021-09-13 18:08:28 +02:00
wtxn.commit().unwrap();
index
},
move |index| {
let config = IndexerConfig::default();
let indexing_config = IndexDocumentsConfig::default();
2021-09-13 18:08:28 +02:00
let mut wtxn = index.write_txn().unwrap();
let mut builder =
IndexDocuments::new(&mut wtxn, &index, &config, indexing_config, |_| ());
let documents = utils::documents_from(datasets_paths::SMOL_ALL_COUNTRIES, "jsonl");
builder.add_documents(documents).unwrap();
builder.execute().unwrap();
2021-09-13 18:08:28 +02:00
wtxn.commit().unwrap();
index.prepare_for_closing().wait();
},
)
});
}
2021-07-07 11:42:14 +02:00
criterion_group!(
benches,
indexing_songs_default,
indexing_songs_without_faceted_numbers,
indexing_songs_without_faceted_fields,
indexing_songs_in_three_batches_default,
2021-07-07 11:42:14 +02:00
indexing_wiki,
indexing_wiki_in_three_batches,
2021-09-13 18:08:28 +02:00
indexing_movies_default,
indexing_movies_in_three_batches,
2021-09-13 18:08:28 +02:00
indexing_geo
2021-07-07 11:42:14 +02:00
);
criterion_main!(benches);