Access term dictionary more efficiently (#92269)

With this commit we sort all ids that are used in `mget` calls in 
lexicographically ascending order. As the term dictionary (which is used to 
lookup `_id`) is also sorted lexicographically, we minimize the number of 
blocks that need to read, reducing page faults.

docs/changelog/92269.yaml

@@ -0,0 +1,5 @@
+pr: 92269
+summary: Access term dictionary more efficiently
+area: Search
+type: enhancement
+issues: []

x-pack/plugin/profiler/src/main/java/org/elasticsearch/xpack/profiler/TransportGetProfilingAction.java

@@ -18,7 +18,6 @@ import org.elasticsearch.client.internal.Client;
 import org.elasticsearch.client.internal.ParentTaskAssigningClient;
 import org.elasticsearch.client.internal.node.NodeClient;
 import org.elasticsearch.common.inject.Inject;
-import org.elasticsearch.common.util.Maps;
 import org.elasticsearch.index.IndexNotFoundException;
 import org.elasticsearch.search.aggregations.bucket.terms.StringTerms;
 import org.elasticsearch.search.aggregations.bucket.terms.TermsAggregationBuilder;
@@ -30,10 +29,11 @@ import org.elasticsearch.xcontent.ObjectPath;
 
 import java.util.Arrays;
 import java.util.HashMap;
-import java.util.HashSet;
 import java.util.Map;
 import java.util.Random;
 import java.util.Set;
+import java.util.TreeMap;
+import java.util.TreeSet;
 
 public class TransportGetProfilingAction extends HandledTransportAction<GetProfilingRequest, GetProfilingResponse> {
     private static final Logger log = LogManager.getLogger(TransportGetProfilingAction.class);
@@ -108,7 +108,10 @@ public class TransportGetProfilingAction extends HandledTransportAction<GetProfi
                     long totalCount = Math.round(totalCountAgg.value());
                     Resampler resampler = new Resampler(request, eventsIndex.getSampleRate(), totalCount);
                     StringTerms stacktraces = searchResponse.getAggregations().get("group_by");
-                    Map<String, Integer> stackTraceEvents = Maps.newHashMapWithExpectedSize(stacktraces.getBuckets().size());
+                    // sort items lexicographically to access Lucene's term dictionary more efficiently when issuing an mget request.
+                    // The term dictionary is lexicographically sorted and using the same order reduces the number of page faults
+                    // needed to load it.
+                    Map<String, Integer> stackTraceEvents = new TreeMap<>();
                     for (StringTerms.Bucket bucket : stacktraces.getBuckets()) {
                         Sum count = bucket.getAggregations().get("count");
                         int finalCount = resampler.adjustSampleCount((int) count.value());
@@ -143,8 +146,11 @@ public class TransportGetProfilingAction extends HandledTransportAction<GetProfi
                 @Override
                 public void onResponse(MultiGetResponse multiGetItemResponses) {
                     Map<String, StackTrace> stackTracePerId = new HashMap<>();
-                    Set<String> stackFrameIds = new HashSet<>();
-                    Set<String> executableIds = new HashSet<>();
+                    // sort items lexicographically to access Lucene's term dictionary more efficiently when issuing an mget request.
+                    // The term dictionary is lexicographically sorted and using the same order reduces the number of page faults
+                    // needed to load it.
+                    Set<String> stackFrameIds = new TreeSet<>();
+                    Set<String> executableIds = new TreeSet<>();
                     int totalFrames = 0;
                     for (MultiGetItemResponse trace : multiGetItemResponses) {
                         if (trace.isFailed() == false && trace.getResponse().isExists()) {