Add a benchmark for transactions retrieval

It consist of two rounds with different db size (respectively 1 000 000
and 10 000 000 transactions) to check that the db size doesn't matter.
Each round will retrieve 100, 10 000, 100 000, and 1 000 000
transactions.

Author: Alenar

mithril-aggregator/Cargo.toml

@@ -13,6 +13,10 @@ repository = { workspace = true }
 name = "cardano_transactions_import"
 harness = false
 
+[[bench]]
+name = "cardano_transactions_get"
+harness = false
+
 [dependencies]
 anyhow = "1.0.79"
 async-trait = "0.1.77"

mithril-aggregator/benches/cardano_transactions_get.rs

@@ -0,0 +1,95 @@
+use std::sync::Arc;
+
+use criterion::{criterion_group, criterion_main, BenchmarkId, Criterion};
+use sqlite::ConnectionThreadSafe;
+
+use mithril_aggregator::{
+    database::repository::CardanoTransactionRepository, services::TransactionStore,
+};
+use mithril_common::{entities::CardanoTransaction, test_utils::TempDir};
+use mithril_persistence::sqlite::ConnectionBuilder;
+
+fn cardano_tx_db_connection(db_file_name: &str) -> ConnectionThreadSafe {
+    let db_path =
+        TempDir::create("aggregator_benches", "bench_get_transactions").join(db_file_name);
+
+    if db_path.exists() {
+        std::fs::remove_file(db_path.clone()).unwrap();
+    }
+
+    ConnectionBuilder::open_file(&db_path)
+        .with_migrations(
+            mithril_aggregator::database::cardano_transaction_migration::get_migrations(),
+        )
+        .build()
+        .unwrap()
+}
+
+fn generate_transactions(nb_transactions: usize) -> Vec<CardanoTransaction> {
+    // Note: we irrealistically generate transactions where each are on a different block.
+    // This is to trick the repository `get_transactions_in_range` method to read the expected number
+    // of transactions.
+    (0..nb_transactions)
+        .map(|i| {
+            CardanoTransaction::new(
+                format!("tx_hash-{}", i),
+                i as u64,
+                i as u64 * 100,
+                format!("block_hash-{}", i),
+                i as u64 + 1,
+            )
+        })
+        .collect()
+}
+
+async fn init_db(nb_transaction_in_db: usize) -> CardanoTransactionRepository {
+    println!("Generating a db with {nb_transaction_in_db} transactions, one per block ...");
+    let transactions = generate_transactions(nb_transaction_in_db);
+    let connection = Arc::new(cardano_tx_db_connection(&format!(
+        "cardano_tx-{nb_transaction_in_db}.db",
+    )));
+    let repository = CardanoTransactionRepository::new(connection);
+    repository.store_transactions(transactions).await.unwrap();
+
+    repository
+}
+
+fn run_bench(c: &mut Criterion, nb_transaction_in_db: usize) {
+    let runtime = tokio::runtime::Runtime::new().unwrap();
+    let repository = runtime.block_on(async { init_db(nb_transaction_in_db).await });
+
+    let mut group = c.benchmark_group(format!(
+        "Get transactions - {nb_transaction_in_db} tx in db"
+    ));
+    for max_block_number in [100, 10_000, 100_000, 1_000_000] {
+        group.bench_with_input(
+            BenchmarkId::from_parameter(format!(
+                "get_transactions_in_range(0..{max_block_number})"
+            )),
+            &max_block_number,
+            |b, &max_block_number| {
+                b.to_async(&runtime).iter(|| async {
+                    let _transactions = repository
+                        .get_transactions_in_range(0..max_block_number)
+                        .await
+                        .unwrap();
+                });
+            },
+        );
+    }
+    group.finish();
+}
+
+fn bench_get_transactions(c: &mut Criterion) {
+    // Two rounds of benchmarks: one with 1M transactions in the db, and one with 10M transactions.
+    // Each time the number of transactions to read is 100, 10_000, 100_000, 1_000_000.
+    run_bench(c, 1_000_000);
+    run_bench(c, 10_000_000);
+}
+
+criterion_group! {
+    name = benches;
+    config = Criterion::default().sample_size(20);
+    targets = bench_get_transactions
+}
+criterion_main!(benches);