Plug the grenad max memory parameter

crates/milli/src/update/index_documents/helpers/grenad_helpers.rs

@@ -119,12 +119,8 @@ impl GrenadParameters {
     ///
     /// This should be called inside of a rayon thread pool,
     /// otherwise, it will take the global number of threads.
-    ///
-    /// The max memory cannot exceed a given reasonable value.
     pub fn max_memory_by_thread(&self) -> Option<usize> {
-        self.max_memory.map(|max_memory| {
-            (max_memory / rayon::current_num_threads()).min(MAX_GRENAD_SORTER_USAGE)
-        })
+        self.max_memory.map(|max_memory| (max_memory / rayon::current_num_threads()))
     }
 }
 

crates/milli/src/update/indexer_config.rs

@@ -1,5 +1,6 @@
 use grenad::CompressionType;
 
+use super::GrenadParameters;
 use crate::thread_pool_no_abort::ThreadPoolNoAbort;
 
 #[derive(Debug)]
@@ -15,6 +16,17 @@ pub struct IndexerConfig {
     pub skip_index_budget: bool,
 }
 
+impl IndexerConfig {
+    pub fn grenad_parameters(&self) -> GrenadParameters {
+        GrenadParameters {
+            chunk_compression_type: self.chunk_compression_type,
+            chunk_compression_level: self.chunk_compression_level,
+            max_memory: self.max_memory,
+            max_nb_chunks: self.max_nb_chunks,
+        }
+    }
+}
+
 impl Default for IndexerConfig {
     fn default() -> Self {
         Self {

crates/milli/src/update/new/extract/faceted/extract_facets.rs

@@ -36,7 +36,7 @@ impl<'a, 'extractor> Extractor<'extractor> for FacetedExtractorData<'a> {
     fn init_data(&self, extractor_alloc: &'extractor Bump) -> Result<Self::Data> {
         Ok(RefCell::new(BalancedCaches::new_in(
             self.buckets,
-            self.grenad_parameters.max_memory,
+            self.grenad_parameters.max_memory_by_thread(),
             extractor_alloc,
         )))
     }

crates/milli/src/update/new/extract/geo/mod.rs

@@ -150,7 +150,7 @@ impl<'extractor> Extractor<'extractor> for GeoExtractor {
     ) -> Result<()> {
         let rtxn = &context.rtxn;
         let index = context.index;
-        let max_memory = self.grenad_parameters.max_memory;
+        let max_memory = self.grenad_parameters.max_memory_by_thread();
         let db_fields_ids_map = context.db_fields_ids_map;
         let mut data_ref = context.data.borrow_mut_or_yield();
 

crates/milli/src/update/new/extract/searchable/extract_word_docids.rs

@@ -214,7 +214,7 @@ impl<'a, 'extractor> Extractor<'extractor> for WordDocidsExtractorData<'a> {
     fn init_data(&self, extractor_alloc: &'extractor Bump) -> Result<Self::Data> {
         Ok(RefCell::new(Some(WordDocidsBalancedCaches::new_in(
             self.buckets,
-            self.grenad_parameters.max_memory,
+            self.grenad_parameters.max_memory_by_thread(),
             extractor_alloc,
         ))))
     }

crates/milli/src/update/new/extract/searchable/mod.rs

@@ -36,7 +36,7 @@ impl<'a, 'extractor, EX: SearchableExtractor + Sync> Extractor<'extractor>
     fn init_data(&self, extractor_alloc: &'extractor Bump) -> Result<Self::Data> {
         Ok(RefCell::new(BalancedCaches::new_in(
             self.buckets,
-            self.grenad_parameters.max_memory,
+            self.grenad_parameters.max_memory_by_thread(),
             extractor_alloc,
         )))
     }

crates/milli/src/update/new/indexer/mod.rs

@@ -132,6 +132,7 @@ mod steps {
 pub fn index<'pl, 'indexer, 'index, DC, MSP, SP>(
     wtxn: &mut RwTxn,
     index: &'index Index,
+    grenad_parameters: GrenadParameters,
     db_fields_ids_map: &'indexer FieldsIdsMap,
     new_fields_ids_map: FieldsIdsMap,
     new_primary_key: Option<PrimaryKey<'pl>>,
@@ -209,16 +210,6 @@ where
 
             field_distribution.retain(|_, v| *v != 0);
 
-            const TEN_GIB: usize = 10 * 1024 * 1024 * 1024;
-            let current_num_threads = rayon::current_num_threads();
-            let max_memory = TEN_GIB / current_num_threads;
-            eprintln!("A maximum of {max_memory} bytes will be used for each of the {current_num_threads} threads");
-
-            let grenad_parameters = GrenadParameters {
-                max_memory: Some(max_memory),
-                ..GrenadParameters::default()
-            };
-
             let facet_field_ids_delta;
 
             {
@@ -228,7 +219,8 @@ where
                 let (finished_steps, step_name) = steps::extract_facets();
 
                 facet_field_ids_delta = merge_and_send_facet_docids(
-                    FacetedDocidsExtractor::run_extraction(grenad_parameters,
+                    FacetedDocidsExtractor::run_extraction(
+                        grenad_parameters,
                         document_changes,
                         indexing_context,
                         &mut extractor_allocs,
@@ -344,7 +336,8 @@ where
 
                 let (finished_steps, step_name) = steps::extract_word_proximity();
 
-                let caches = <WordPairProximityDocidsExtractor as DocidsExtractor>::run_extraction(grenad_parameters,
+                let caches = <WordPairProximityDocidsExtractor as DocidsExtractor>::run_extraction(
+                    grenad_parameters,
                     document_changes,
                     indexing_context,
                     &mut extractor_allocs,
@@ -398,7 +391,8 @@ where
                 };
                 let datastore = ThreadLocal::with_capacity(rayon::current_num_threads());
                 let (finished_steps, step_name) = steps::extract_geo_points();
-                extract(document_changes,
+                extract(
+                    document_changes,
                     &extractor,
                     indexing_context,
                     &mut extractor_allocs,