Merge branch 'main' of github.com:samoii/quickwit into aws-s3-sse

docs/reference/es_compatible_api.md

@@ -135,6 +135,7 @@ If a parameter appears both as a query string parameter and in the JSON payload,
 | `size`             | `Integer`     | Number of hits to return.                                                        | 10            |
 | `sort`             | `String`      | Describes how documents should be ranked. See [Sort order](#sort-order)          | (Optional)    |
 | `scroll`           | `Duration`    | Creates a scroll context for "time to live". See [Scroll](#_scroll--scroll-api). | (Optional)    |
+| `allow_partial_search_results` | `Boolean` | Returns a partial response if some (but not all) of the split searches were unsuccessful. | `true` |
 
 #### Supported Request Body parameters
 
@@ -301,7 +302,7 @@ GET api/v1/_elastic/_cat/indices
 
 Use the [cat indices API](https://www.elastic.co/guide/en/elasticsearch/reference/current/cat-indices.html) to get the following information for each index in a cluster:
 * Shard count
-* Document count 
+* Document count
 * Deleted document count
 * Primary store size
 * Total store size

monitoring/grafana/dashboards/searchers.json

@@ -853,6 +853,10 @@
           "text": "All",
           "value": "$__all"
         },
+        "datasource": {
+          "type": "prometheus",
+          "uid": "${datasource}"
+        },
         "definition": "label_values(quickwit_search_leaf_searches_splits_total,instance)",
         "hide": 0,
         "includeAll": true,
@@ -861,16 +865,16 @@
         "name": "instance",
         "options": [],
         "query": {
-          "qryType": 1,
           "query": "label_values(quickwit_search_leaf_searches_splits_total,instance)",
-          "refId": "PrometheusVariableQueryEditor-VariableQuery"
+          "refId": "StandardVariableQuery"
         },
         "refresh": 1,
         "regex": "",
         "skipUrlSync": false,
         "sort": 0,
         "type": "query"
       }
+
     ]
   },
   "time": {

quickwit/quickwit-cli/src/tool.rs

@@ -550,6 +550,7 @@ pub async fn local_search_cli(args: LocalSearchArgs) -> anyhow::Result<()> {
         format: BodyFormat::Json,
         sort_by,
         count_all: CountHits::CountAll,
+        allow_failed_splits: false,
     };
     let search_request =
         search_request_from_api_request(vec![args.index_id], search_request_query_string)?;

quickwit/quickwit-proto/protos/quickwit/search.proto

@@ -126,14 +126,14 @@ message ListFieldsRequest {
   // Optional limit query to a list of fields
   // Wildcard expressions are supported.
   repeated string fields = 2;
-  
+
   // Time filter, expressed in seconds since epoch.
   // That filter is to be interpreted as the semi-open interval:
   // [start_timestamp, end_timestamp).
   optional int64 start_timestamp = 3;
   optional int64 end_timestamp = 4;
 
-  // Control if the the request will fail if split_ids contains a split that does not exist. 
+  // Control if the the request will fail if split_ids contains a split that does not exist.
   // optional bool fail_on_missing_index = 6;
 }
 
@@ -149,7 +149,7 @@ message LeafListFieldsRequest {
   // Optional limit query to a list of fields
   // Wildcard expressions are supported.
   repeated string fields = 4;
-  
+
 }
 
 message ListFieldsResponse {
@@ -299,6 +299,17 @@ message SearchResponse {
 
   // Scroll Id (only set if scroll_secs was set in the request)
   optional string scroll_id = 6;
+
+  // Returns the list of splits for which search failed.
+  // For the moment, the cause is unknown.
+  //
+  // It is up to the caller to decide whether to interpret
+  // this as an overall failure or to present the partial results
+  // to the end user.
+  repeated SplitSearchError failed_splits = 7;
+
+  // Total number of successful splits searched.
+  uint64 num_successful_splits = 8;
 }
 
 message SearchPlanResponse {
@@ -340,7 +351,7 @@ message LeafSearchRequest {
 message LeafRequestRef {
   // The ordinal of the doc_mapper in `LeafSearchRequest.doc_mappers`
   uint32 doc_mapper_ord = 1;
-  
+
   // The ordinal of the index uri in LeafSearchRequest.index_uris
   uint32 index_uri_ord = 2;
 
@@ -453,10 +464,16 @@ message LeafSearchResponse {
   // The list of splits that failed. LeafSearchResponse can be an aggregation of results, so there may be multiple.
   repeated SplitSearchError failed_splits = 3;
 
-  // Total number of splits the leaf(s) were in charge of.
-  // num_attempted_splits = num_successful_splits + num_failed_splits.
+  // Total number of attempt to search into splits.
+  // We do have:
+  // `num_splits_requested == num_successful_splits + num_failed_splits.len()`
+  // But we do not necessarily have:
+  // `num_splits_requested = num_attempted_splits because of retries.`
   uint64 num_attempted_splits = 4;
 
+  // Total number of successful splits searched.
+  uint64 num_successful_splits = 7;
+
   // Deprecated json serialized intermediate aggregation_result.
   reserved 5;
 
@@ -550,8 +567,7 @@ message LeafListTermsResponse {
   // The list of splits that failed. LeafSearchResponse can be an aggregation of results, so there may be multiple.
   repeated SplitSearchError failed_splits = 3;
 
-  // Total number of splits the leaf(s) were in charge of.
-  // num_attempted_splits = num_successful_splits + num_failed_splits.
+  // Total number of single split search attempted.
   uint64 num_attempted_splits = 4;
 }
 

quickwit/quickwit-search/src/client.rs

@@ -20,7 +20,7 @@
 use std::fmt;
 use std::net::SocketAddr;
 use std::sync::Arc;
-use std::time::Duration;
+use std::time::{Duration, Instant};
 
 use bytesize::ByteSize;
 use futures::{StreamExt, TryStreamExt};
@@ -37,6 +37,7 @@ use tower::timeout::Timeout;
 use tracing::{info_span, warn, Instrument};
 
 use crate::error::parse_grpc_error;
+use crate::metrics::SEARCH_METRICS;
 use crate::SearchService;
 
 /// Impl is an enumeration that meant to manage Quickwit's search service client types.
@@ -110,35 +111,61 @@ impl SearchServiceClient {
         &mut self,
         request: quickwit_proto::search::SearchRequest,
     ) -> crate::Result<quickwit_proto::search::SearchResponse> {
-        match &mut self.client_impl {
-            SearchServiceClientImpl::Grpc(grpc_client) => {
-                let tonic_request = Request::new(request);
-                let tonic_response = grpc_client
-                    .root_search(tonic_request)
-                    .await
-                    .map_err(|tonic_error| parse_grpc_error(&tonic_error))?;
-                Ok(tonic_response.into_inner())
-            }
+        let start = Instant::now();
+        let response_result = match &mut self.client_impl {
+            SearchServiceClientImpl::Grpc(grpc_client) => grpc_client
+                .root_search(request)
+                .await
+                .map(|tonic_response| tonic_response.into_inner())
+                .map_err(|tonic_error| parse_grpc_error(&tonic_error)),
             SearchServiceClientImpl::Local(service) => service.root_search(request).await,
-        }
+        };
+        let elapsed = start.elapsed().as_secs_f64();
+        let label_values = if response_result.is_ok() {
+            ["success"]
+        } else {
+            ["error"]
+        };
+        SEARCH_METRICS
+            .root_search_requests_total
+            .with_label_values(label_values)
+            .inc();
+        SEARCH_METRICS
+            .root_search_request_duration_seconds
+            .with_label_values(label_values)
+            .observe(elapsed);
+        response_result
     }
 
     /// Perform leaf search.
     pub async fn leaf_search(
         &mut self,
         request: quickwit_proto::search::LeafSearchRequest,
     ) -> crate::Result<quickwit_proto::search::LeafSearchResponse> {
-        match &mut self.client_impl {
-            SearchServiceClientImpl::Grpc(grpc_client) => {
-                let tonic_request = Request::new(request);
-                let tonic_response = grpc_client
-                    .leaf_search(tonic_request)
-                    .await
-                    .map_err(|tonic_error| parse_grpc_error(&tonic_error))?;
-                Ok(tonic_response.into_inner())
-            }
+        let start = Instant::now();
+        let response_result = match &mut self.client_impl {
+            SearchServiceClientImpl::Grpc(grpc_client) => grpc_client
+                .leaf_search(request)
+                .await
+                .map(|tonic_response| tonic_response.into_inner())
+                .map_err(|tonic_error| parse_grpc_error(&tonic_error)),
             SearchServiceClientImpl::Local(service) => service.leaf_search(request).await,
-        }
+        };
+        let elapsed = start.elapsed().as_secs_f64();
+        let label_values = if response_result.is_ok() {
+            ["success"]
+        } else {
+            ["error"]
+        };
+        SEARCH_METRICS
+            .leaf_search_requests_total
+            .with_label_values(label_values)
+            .inc();
+        SEARCH_METRICS
+            .leaf_search_request_duration_seconds
+            .with_label_values(label_values)
+            .observe(elapsed);
+        response_result
     }
 
     /// Perform leaf search.