improve query performance for samples

EBI-Metagenomics · Dec 6, 2017 · cf1bf68 · cf1bf68
1 parent a89de97
commit cf1bf68
Show file tree

Hide file tree

Showing 3 changed files with 14 additions and 11 deletions.
diff --git a/emgapi/models.py b/emgapi/models.py
@@ -419,17 +419,22 @@ def get_queryset(self):
     def available(self, request, prefetch=False):
         queryset = self.get_queryset().available(request)
         if prefetch:
-            _qs = SampleAnn.objects.all()
             queryset = queryset.extra(
                 {
                     'longitude': "CAST(longitude as DECIMAL(10,5))",
                     'latitude': "CAST(latitude as DECIMAL(10,5))"
                 }
-            )
-            queryset = queryset.prefetch_related(
+            ).prefetch_related(
                 Prefetch('biome', queryset=Biome.objects.all()),
                 Prefetch('studies', queryset=Study.objects.available(request)),
-                Prefetch('metadata', queryset=_qs)
+                Prefetch('metadata', queryset=SampleAnn.objects.all())
+            ).defer(
+                'primary_accession',
+                'is_public',
+                'metadata_received',
+                'sequencedata_received',
+                'sequencedata_archived',
+                'submission_account_id',
             )
         return queryset
 
@@ -829,9 +834,7 @@ class SampleAnnManager(models.Manager):
 
     def get_queryset(self):
         return SampleAnnQuerySet(self.model, using=self._db) \
-            .prefetch_related(
-                Prefetch('var', queryset=VariableNames.objects.all())
-            )
+            .select_related('var')
 
 
 class SampleAnn(models.Model):

diff --git a/emgapi/serializers.py b/emgapi/serializers.py
@@ -684,7 +684,7 @@ def get_biosample(self, obj):
     )
 
     def get_studies(self, obj):
-        return obj.studies.all()
+        return obj.studies.available(self.context['request'])
 
     runs = relations.SerializerMethodResourceRelatedField(
         source='get_runs',
@@ -757,7 +757,9 @@ def get_bioproject(self, obj):
     )
 
     def get_biomes(self, obj):
-        biomes = obj.samples.values('biome_id').distinct()
+        biomes = obj.samples \
+            .available(self.context['request']) \
+            .values('biome_id').distinct()
         return emg_models.Biome.objects \
             .filter(pk__in=biomes)
 

diff --git a/emgapi/views_relations.py b/emgapi/views_relations.py
@@ -111,8 +111,6 @@ def get_queryset(self):
             .available(self.request, prefetch=True) \
             .filter(studies__in=[study])
         _qs = emg_models.Study.objects.available(self.request)
-        # queryset = queryset.prefetch_related(
-        #     Prefetch('studies', queryset=_qs))
         if 'runs' in self.request.GET.get('include', '').split(','):
             _qs = emg_models.Run.objects \
                 .available(self.request)