Merge 94a1f5a8ea into 94fb55bb6f

5049: Fix the path used in the flaky tests CI r=irevoire a=Kerollmops

This PR fixes [the flaky tests CI](https://github.com/meilisearch/meilisearch/actions/runs/11741717787) path used.

Co-authored-by: Clément Renault <clement@meilisearch.com>

.github/workflows/flaky-tests.yml

@@ -21,10 +21,10 @@ jobs:
     - name: Install cargo-flaky
       run: cargo install cargo-flaky
     - name: Run cargo flaky in the dumps
-      run: cd dump; cargo flaky -i 100 --release
+      run: cd crates/dump; cargo flaky -i 100 --release
     - name: Run cargo flaky in the index-scheduler
-      run: cd index-scheduler; cargo flaky -i 100 --release
+      run: cd crates/index-scheduler; cargo flaky -i 100 --release
     - name: Run cargo flaky in the auth
-      run: cd meilisearch-auth; cargo flaky -i 100 --release
+      run: cd crates/meilisearch-auth; cargo flaky -i 100 --release
     - name: Run cargo flaky in meilisearch
-      run: cd meilisearch; cargo flaky -i 100 --release
+      run: cd crates/meilisearch; cargo flaky -i 100 --release

.github/workflows/test-suite.yml

@@ -33,7 +33,7 @@ jobs:
       - name: Setup test with Rust stable
         uses: dtolnay/rust-toolchain@1.79
       - name: Cache dependencies
-        uses: Swatinem/rust-cache@v2.7.1
+        uses: Swatinem/rust-cache@v2.7.5
       - name: Run cargo check without any default features
         uses: actions-rs/cargo@v1
         with:
@@ -55,7 +55,7 @@ jobs:
     steps:
       - uses: actions/checkout@v3
       - name: Cache dependencies
-        uses: Swatinem/rust-cache@v2.7.1
+        uses: Swatinem/rust-cache@v2.7.5
       - uses: dtolnay/rust-toolchain@1.79
       - name: Run cargo check without any default features
         uses: actions-rs/cargo@v1
@@ -127,7 +127,7 @@ jobs:
           apt-get install build-essential -y
       - uses: dtolnay/rust-toolchain@1.79
       - name: Cache dependencies
-        uses: Swatinem/rust-cache@v2.7.1
+        uses: Swatinem/rust-cache@v2.7.5
       - name: Run tests in debug
         uses: actions-rs/cargo@v1
         with:
@@ -144,7 +144,7 @@ jobs:
           profile: minimal
           components: clippy
       - name: Cache dependencies
-        uses: Swatinem/rust-cache@v2.7.1
+        uses: Swatinem/rust-cache@v2.7.5
       - name: Run cargo clippy
         uses: actions-rs/cargo@v1
         with:
@@ -163,7 +163,7 @@ jobs:
           override: true
           components: rustfmt
       - name: Cache dependencies
-        uses: Swatinem/rust-cache@v2.7.1
+        uses: Swatinem/rust-cache@v2.7.5
       - name: Run cargo fmt
         # Since we never ran the `build.rs` script in the benchmark directory we are missing one auto-generated import file.
         # Since we want to trigger (and fail) this action as fast as possible, instead of building the benchmark crate

BENCHMARKS.md

@@ -48,6 +48,27 @@ cargo xtask bench --no-dashboard -- workloads/my_workload_1.json workloads/my_wo
 
 For processing the results, look at [Looking at benchmark results/Without dashboard](#without-dashboard).
 
+#### Sending a workload by hand
+
+Sometimes you want to visualize the metrics of a worlkoad that comes from a custom report.
+It is not quite easy to trick the benchboard in thinking that your report is legitimate but here are the commands you can run to upload your firefox report on a running benchboard.
+
+```bash
+# Name this hostname whatever you want
+echo '{ "hostname": "the-best-place" }' | xh PUT 'http://127.0.0.1:9001/api/v1/machine'
+
+# You'll receive an UUID from this command that we will call $invocation_uuid
+echo '{ "commit": { "sha1": "1234567", "commit_date": "2024-09-05 12:00:12.0 +00:00:00", "message": "A cool message" }, "machine_hostname": "the-best-place", "max_workloads": 1 }' | xh PUT 'http://127.0.0.1:9001/api/v1/invocation'
+
+# Just use UUID from the previous command
+# and you'll receive another UUID that we will call $workload_uuid
+echo '{ "invocation_uuid": "$invocation_uuid", "name": "toto", "max_runs": 1 }' | xh PUT 'http://127.0.0.1:9001/api/v1/workload'
+
+# And now use your $workload_uuid and the content of your firefox report
+# but don't forget to convert your firefox report from JSONLines into an object
+echo '{ "workload_uuid": "$workload_uuid", "data": $REPORT_JSON_DATA }' | xh PUT 'http://127.0.0.1:9001/api/v1/run'
+```
+
 ### In CI
 
 We have dedicated runners to run workloads on CI. Currently, there are three ways of running the CI:

crates/meilisearch/src/metrics.rs

@@ -49,4 +49,18 @@ lazy_static! {
     pub static ref MEILISEARCH_IS_INDEXING: IntGauge =
         register_int_gauge!(opts!("meilisearch_is_indexing", "Meilisearch Is Indexing"))
             .expect("Can't create a metric");
+    pub static ref MEILISEARCH_SEARCH_QUEUE_SIZE: IntGauge = register_int_gauge!(opts!(
+        "meilisearch_search_queue_size",
+        "Meilisearch Search Queue Size"
+    ))
+    .expect("Can't create a metric");
+    pub static ref MEILISEARCH_SEARCHES_RUNNING: IntGauge =
+        register_int_gauge!(opts!("meilisearch_searches_running", "Meilisearch Searches Running"))
+            .expect("Can't create a metric");
+    pub static ref MEILISEARCH_SEARCHES_WAITING_TO_BE_PROCESSED: IntGauge =
+        register_int_gauge!(opts!(
+            "meilisearch_searches_waiting_to_be_processed",
+            "Meilisearch Searches Being Processed"
+        ))
+        .expect("Can't create a metric");
 }

crates/meilisearch/src/routes/metrics.rs

@@ -10,6 +10,7 @@ use prometheus::{Encoder, TextEncoder};
 use crate::extractors::authentication::policies::ActionPolicy;
 use crate::extractors::authentication::{AuthenticationError, GuardedData};
 use crate::routes::create_all_stats;
+use crate::search_queue::SearchQueue;
 
 pub fn configure(config: &mut web::ServiceConfig) {
     config.service(web::resource("").route(web::get().to(get_metrics)));
@@ -18,6 +19,7 @@ pub fn configure(config: &mut web::ServiceConfig) {
 pub async fn get_metrics(
     index_scheduler: GuardedData<ActionPolicy<{ actions::METRICS_GET }>, Data<IndexScheduler>>,
     auth_controller: Data<AuthController>,
+    search_queue: web::Data<SearchQueue>,
 ) -> Result<HttpResponse, ResponseError> {
     index_scheduler.features().check_metrics()?;
     let auth_filters = index_scheduler.filters();
@@ -35,6 +37,11 @@ pub async fn get_metrics(
     crate::metrics::MEILISEARCH_USED_DB_SIZE_BYTES.set(response.used_database_size as i64);
     crate::metrics::MEILISEARCH_INDEX_COUNT.set(response.indexes.len() as i64);
 
+    crate::metrics::MEILISEARCH_SEARCH_QUEUE_SIZE.set(search_queue.capacity() as i64);
+    crate::metrics::MEILISEARCH_SEARCHES_RUNNING.set(search_queue.searches_running() as i64);
+    crate::metrics::MEILISEARCH_SEARCHES_WAITING_TO_BE_PROCESSED
+        .set(search_queue.searches_waiting() as i64);
+
     for (index, value) in response.indexes.iter() {
         crate::metrics::MEILISEARCH_INDEX_DOCS_COUNT
             .with_label_values(&[index])

crates/meilisearch/src/search_queue.rs

@@ -18,6 +18,8 @@
 //!                         And should drop the Permit only once you have freed all the RAM consumed by the method.
 
 use std::num::NonZeroUsize;
+use std::sync::atomic::{AtomicUsize, Ordering};
+use std::sync::Arc;
 use std::time::Duration;
 
 use rand::rngs::StdRng;
@@ -33,6 +35,8 @@ pub struct SearchQueue {
     /// If we have waited longer than this to get a permit, we should abort the search request entirely.
     /// The client probably already closed the connection, but we have no way to find out.
     time_to_abort: Duration,
+    searches_running: Arc<AtomicUsize>,
+    searches_waiting_to_be_processed: Arc<AtomicUsize>,
 }
 
 /// You should only run search requests while holding this permit.
@@ -68,14 +72,41 @@ impl SearchQueue {
         // so let's not allocate any RAM and keep a capacity of 1.
         let (sender, receiver) = mpsc::channel(1);
 
-        tokio::task::spawn(Self::run(capacity, paralellism, receiver));
-        Self { sender, capacity, time_to_abort: Duration::from_secs(60) }
+        let instance = Self {
+            sender,
+            capacity,
+            time_to_abort: Duration::from_secs(60),
+            searches_running: Default::default(),
+            searches_waiting_to_be_processed: Default::default(),
+        };
+
+        tokio::task::spawn(Self::run(
+            capacity,
+            paralellism,
+            receiver,
+            Arc::clone(&instance.searches_running),
+            Arc::clone(&instance.searches_waiting_to_be_processed),
+        ));
+
+        instance
     }
 
     pub fn with_time_to_abort(self, time_to_abort: Duration) -> Self {
         Self { time_to_abort, ..self }
     }
 
+    pub fn capacity(&self) -> usize {
+        self.capacity
+    }
+
+    pub fn searches_running(&self) -> usize {
+        self.searches_running.load(Ordering::Relaxed)
+    }
+
+    pub fn searches_waiting(&self) -> usize {
+        self.searches_waiting_to_be_processed.load(Ordering::Relaxed)
+    }
+
     /// This function is the main loop, it's in charge on scheduling which search request should execute first and
     /// how many should executes at the same time.
     ///
@@ -84,6 +115,8 @@ impl SearchQueue {
         capacity: usize,
         parallelism: NonZeroUsize,
         mut receive_new_searches: mpsc::Receiver<oneshot::Sender<Permit>>,
+        metric_searches_running: Arc<AtomicUsize>,
+        metric_searches_waiting: Arc<AtomicUsize>,
     ) {
         let mut queue: Vec<oneshot::Sender<Permit>> = Default::default();
         let mut rng: StdRng = StdRng::from_entropy();
@@ -133,6 +166,9 @@ impl SearchQueue {
                     queue.push(search_request);
                 },
             }
+
+            metric_searches_running.store(searches_running, Ordering::Relaxed);
+            metric_searches_waiting.store(queue.len(), Ordering::Relaxed);
         }
     }
 

crates/meilisearch/tests/common/mod.rs

@@ -389,3 +389,25 @@ pub static VECTOR_DOCUMENTS: Lazy<Value> = Lazy::new(|| {
       },
     ])
 });
+
+pub async fn shared_index_with_test_set() -> &'static Index<'static, Shared> {
+    static INDEX: OnceCell<Index<'static, Shared>> = OnceCell::const_new();
+    INDEX
+        .get_or_init(|| async {
+            let server = Server::new_shared();
+            let index = server._index("SHARED_TEST_SET").to_shared();
+            let url = format!("/indexes/{}/documents", urlencoding::encode(index.uid.as_ref()));
+            let (response, code) = index
+                .service
+                .post_str(
+                    url,
+                    include_str!("../assets/test_set.json"),
+                    vec![("content-type", "application/json")],
+                )
+                .await;
+            assert_eq!(code, 202);
+            index.wait_task(response.uid()).await;
+            index
+        })
+        .await
+}

crates/meilisearch/tests/documents/get_documents.rs

@@ -4,24 +4,27 @@ use meili_snap::*;
 use urlencoding::encode as urlencode;
 
 use crate::common::encoder::Encoder;
-use crate::common::{GetAllDocumentsOptions, Server, Value};
+use crate::common::{
+    shared_does_not_exists_index, shared_empty_index, shared_index_with_test_set,
+    GetAllDocumentsOptions, Server, Value,
+};
 use crate::json;
 
 // TODO: partial test since we are testing error, amd error is not yet fully implemented in
 // transplant
 #[actix_rt::test]
 async fn get_unexisting_index_single_document() {
-    let server = Server::new().await;
-    let (_response, code) = server.index("test").get_document(1, None).await;
+    let (_response, code) = shared_does_not_exists_index().await.get_document(1, None).await;
     assert_eq!(code, 404);
 }
 
 #[actix_rt::test]
 async fn error_get_unexisting_document() {
-    let server = Server::new().await;
-    let index = server.index("test");
-    index.create(None).await;
-    index.wait_task(0).await;
+    let server = Server::new_shared();
+    let index = server.unique_index();
+    let (task, _code) = index.create(None).await;
+    index.wait_task(task.uid()).await.succeeded();
+
     let (response, code) = index.get_document(1, None).await;
 
     let expected_response = json!({
@@ -37,18 +40,19 @@ async fn error_get_unexisting_document() {
 
 #[actix_rt::test]
 async fn get_document() {
-    let server = Server::new().await;
-    let index = server.index("test");
-    index.create(None).await;
+    let server = Server::new_shared();
+    let index = server.unique_index();
+    let (task, _code) = index.create(None).await;
+    index.wait_task(task.uid()).await.succeeded();
     let documents = json!([
         {
             "id": 0,
             "nested": { "content": "foobar" },
         }
     ]);
-    let (_, code) = index.add_documents(documents, None).await;
+    let (task, code) = index.add_documents(documents, None).await;
     assert_eq!(code, 202);
-    index.wait_task(1).await;
+    index.wait_task(task.uid()).await.succeeded();
     let (response, code) = index.get_document(0, None).await;
     assert_eq!(code, 200);
     assert_eq!(
@@ -81,12 +85,11 @@ async fn get_document() {
 
 #[actix_rt::test]
 async fn error_get_unexisting_index_all_documents() {
-    let server = Server::new().await;
-    let (response, code) =
-        server.index("test").get_all_documents(GetAllDocumentsOptions::default()).await;
+    let index = shared_does_not_exists_index().await;
+    let (response, code) = index.get_all_documents(GetAllDocumentsOptions::default()).await;
 
     let expected_response = json!({
-        "message": "Index `test` not found.",
+        "message": "Index `DOES_NOT_EXISTS` not found.",
         "code": "index_not_found",
         "type": "invalid_request",
         "link": "https://docs.meilisearch.com/errors#index_not_found"
@@ -98,12 +101,7 @@ async fn error_get_unexisting_index_all_documents() {
 
 #[actix_rt::test]
 async fn get_no_document() {
-    let server = Server::new().await;
-    let index = server.index("test");
-    let (_, code) = index.create(None).await;
-    assert_eq!(code, 202);
-
-    index.wait_task(0).await;
+    let index = shared_empty_index().await;
 
     let (response, code) = index.get_all_documents(GetAllDocumentsOptions::default()).await;
     assert_eq!(code, 200);
@@ -112,14 +110,12 @@ async fn get_no_document() {
 
 #[actix_rt::test]
 async fn get_all_documents_no_options() {
-    let server = Server::new().await;
-    let index = server.index("test");
-    index.load_test_set().await;
+    let index = shared_index_with_test_set().await;
 
     let (response, code) = index.get_all_documents(GetAllDocumentsOptions::default()).await;
     assert_eq!(code, 200);
-    let arr = response["results"].as_array().unwrap();
-    assert_eq!(arr.len(), 20);
+    let results = response["results"].as_array().unwrap();
+    assert_eq!(results.len(), 20);
     let first = json!({
         "id":0,
         "isActive":false,
@@ -138,19 +134,16 @@ async fn get_all_documents_no_options() {
         "longitude":-145.725388,
         "tags":["bug"
             ,"bug"]});
-    assert_eq!(first, arr[0]);
+    assert_eq!(first, results[0]);
 }
 
 #[actix_rt::test]
 async fn get_all_documents_no_options_with_response_compression() {
-    let server = Server::new().await;
-    let index_uid = "test";
-    let index = server.index(index_uid);
-    index.load_test_set().await;
+    let index = shared_index_with_test_set().await;
 
-    let app = server.init_web_app().await;
+    let app = Server::new_shared().init_web_app().await;
     let req = test::TestRequest::get()
-        .uri(&format!("/indexes/{}/documents?", urlencode(index_uid)))
+        .uri(&format!("/indexes/{}/documents?", urlencode(&index.uid)))
         .insert_header((ACCEPT_ENCODING, "gzip"))
         .to_request();
 
@@ -169,9 +162,7 @@ async fn get_all_documents_no_options_with_response_compression() {
 
 #[actix_rt::test]
 async fn test_get_all_documents_limit() {
-    let server = Server::new().await;
-    let index = server.index("test");
-    index.load_test_set().await;
+    let index = shared_index_with_test_set().await;
 
     let (response, code) = index
         .get_all_documents(GetAllDocumentsOptions { limit: Some(5), ..Default::default() })
@@ -186,9 +177,7 @@ async fn test_get_all_documents_limit() {
 
 #[actix_rt::test]
 async fn test_get_all_documents_offset() {
-    let server = Server::new().await;
-    let index = server.index("test");
-    index.load_test_set().await;
+    let index = shared_index_with_test_set().await;
 
     let (response, code) = index
         .get_all_documents(GetAllDocumentsOptions { offset: Some(5), ..Default::default() })
@@ -203,9 +192,7 @@ async fn test_get_all_documents_offset() {
 
 #[actix_rt::test]
 async fn test_get_all_documents_attributes_to_retrieve() {
-    let server = Server::new().await;
-    let index = server.index("test");
-    index.load_test_set().await;
+    let index = shared_index_with_test_set().await;
 
     let (response, code) = index
         .get_all_documents(GetAllDocumentsOptions {
@@ -286,9 +273,11 @@ async fn test_get_all_documents_attributes_to_retrieve() {
 
 #[actix_rt::test]
 async fn get_document_s_nested_attributes_to_retrieve() {
-    let server = Server::new().await;
-    let index = server.index("test");
-    index.create(None).await;
+    let server = Server::new_shared();
+    let index = server.unique_index();
+    let (task, _code) = index.create(None).await;
+    index.wait_task(task.uid()).await.succeeded();
+
     let documents = json!([
         {
             "id": 0,
@@ -302,9 +291,9 @@ async fn get_document_s_nested_attributes_to_retrieve() {
             },
         },
     ]);
-    let (_, code) = index.add_documents(documents, None).await;
+    let (task, code) = index.add_documents(documents, None).await;
     assert_eq!(code, 202);
-    index.wait_task(1).await;
+    index.wait_task(task.uid()).await.succeeded();
 
     let (response, code) = index.get_document(0, Some(json!({ "fields": ["content"] }))).await;
     assert_eq!(code, 200);
@@ -343,10 +332,10 @@ async fn get_document_s_nested_attributes_to_retrieve() {
 
 #[actix_rt::test]
 async fn get_documents_displayed_attributes_is_ignored() {
-    let server = Server::new().await;
-    let index = server.index("test");
-    index.update_settings(json!({"displayedAttributes": ["gender"]})).await;
+    let server = Server::new_shared();
+    let index = server.unique_index();
     index.load_test_set().await;
+    index.update_settings(json!({"displayedAttributes": ["gender"]})).await;
 
     let (response, code) = index.get_all_documents(GetAllDocumentsOptions::default()).await;
     assert_eq!(code, 200);
@@ -366,10 +355,10 @@ async fn get_documents_displayed_attributes_is_ignored() {
 
 #[actix_rt::test]
 async fn get_document_by_filter() {
-    let server = Server::new().await;
-    let index = server.index("doggo");
+    let server = Server::new_shared();
+    let index = server.unique_index();
     index.update_settings_filterable_attributes(json!(["color"])).await;
-    index
+    let (task, _code) = index
         .add_documents(
             json!([
                 { "id": 0, "color": "red" },
@@ -380,7 +369,7 @@ async fn get_document_by_filter() {
             Some("id"),
         )
         .await;
-    index.wait_task(1).await;
+    index.wait_task(task.uid()).await.succeeded();
 
     let (response, code) = index.get_document_by_filter(json!({})).await;
     let (response2, code2) = index.get_all_documents_raw("").await;
@@ -552,7 +541,7 @@ async fn get_document_with_vectors() {
         }))
         .await;
     snapshot!(code, @"202 Accepted");
-    server.wait_task(response.uid()).await;
+    server.wait_task(response.uid()).await.succeeded();
 
     let documents = json!([
       {"id": 0, "name": "kefir", "_vectors": { "manual": [0, 0, 0] }},
@@ -560,7 +549,7 @@ async fn get_document_with_vectors() {
     ]);
     let (value, code) = index.add_documents(documents, None).await;
     snapshot!(code, @"202 Accepted");
-    index.wait_task(value.uid()).await;
+    index.wait_task(value.uid()).await.succeeded();
 
     // by default you shouldn't see the `_vectors` object
     let (documents, _code) = index.get_all_documents(Default::default()).await;

crates/meilisearch/tests/search/formatted.rs

@@ -6,14 +6,14 @@ use crate::json;
 
 #[actix_rt::test]
 async fn formatted_contain_wildcard() {
-    let server = Server::new().await;
-    let index = server.index("test");
+    let server = Server::new_shared();
+    let index = server.unique_index();
 
     index.update_settings(json!({ "displayedAttributes": ["id", "cattos"] })).await;
 
     let documents = NESTED_DOCUMENTS.clone();
-    index.add_documents(documents, None).await;
-    index.wait_task(1).await;
+    let (response, _) = index.add_documents(documents, None).await;
+    index.wait_task(response.uid()).await;
 
     index.search(json!({ "q": "pésti", "attributesToRetrieve": ["father", "mother"], "attributesToHighlight": ["father", "mother", "*"], "attributesToCrop": ["doggos"], "showMatchesPosition": true }),
         |response, code|
@@ -135,12 +135,7 @@ async fn formatted_contain_wildcard() {
 
 #[actix_rt::test]
 async fn format_nested() {
-    let server = Server::new().await;
-    let index = server.index("test");
-
-    let documents = NESTED_DOCUMENTS.clone();
-    index.add_documents(documents, None).await;
-    index.wait_task(0).await;
+    let index = shared_index_with_nested_documents().await;
 
     index
         .search(json!({ "q": "pésti", "attributesToRetrieve": ["doggos"] }), |response, code| {
@@ -340,15 +335,15 @@ async fn format_nested() {
 
 #[actix_rt::test]
 async fn displayedattr_2_smol() {
-    let server = Server::new().await;
-    let index = server.index("test");
+    let server = Server::new_shared();
+    let index = server.unique_index();
 
     // not enough displayed for the other settings
     index.update_settings(json!({ "displayedAttributes": ["id"] })).await;
 
     let documents = NESTED_DOCUMENTS.clone();
-    index.add_documents(documents, None).await;
-    index.wait_task(1).await;
+    let (response, _) = index.add_documents(documents, None).await;
+    index.wait_task(response.uid()).await;
 
     index
         .search(json!({ "attributesToRetrieve": ["father", "id"], "attributesToHighlight": ["mother"], "attributesToCrop": ["cattos"] }),
@@ -538,15 +533,15 @@ async fn displayedattr_2_smol() {
 #[cfg(feature = "default")]
 #[actix_rt::test]
 async fn test_cjk_highlight() {
-    let server = Server::new().await;
-    let index = server.index("test");
+    let server = Server::new_shared();
+    let index = server.unique_index();
 
     let documents = json!([
         { "id": 0, "title": "この度、クーポンで無料で頂きました。" },
         { "id": 1, "title": "大卫到了扫罗那里" },
     ]);
-    index.add_documents(documents, None).await;
-    index.wait_task(0).await;
+    let (response, _) = index.add_documents(documents, None).await;
+    index.wait_task(response.uid()).await;
 
     index
         .search(json!({"q": "で", "attributesToHighlight": ["title"]}), |response, code| {