From a2af94ea693e437564d53043a5dbc30d409dc4ac Mon Sep 17 00:00:00 2001
From: John Tordoff <>
Date: Mon, 7 Oct 2024 12:51:12 -0400
Subject: [PATCH] fix up issue with lingering connections and optimize query a
 tad using pre-select

---
 .../reporters/institution_summary_monthly.py        | 13 ++++++-------
 1 file changed, 6 insertions(+), 7 deletions(-)

diff --git a/osf/metrics/reporters/institution_summary_monthly.py b/osf/metrics/reporters/institution_summary_monthly.py
index a46a7af8a98..95baffd0f7b 100644
--- a/osf/metrics/reporters/institution_summary_monthly.py
+++ b/osf/metrics/reporters/institution_summary_monthly.py
@@ -1,6 +1,7 @@
+from django.db import connection
+from django.db.models import Q, F, Sum, Count
+from django.db import close_old_connections
 from django.contrib.contenttypes.models import ContentType
-from django.db.models import Q, F, Sum
-
 from osf.models import Institution, Preprint, AbstractNode, FileVersion
 from osf.models.spam import SpamStatus
 from addons.osfstorage.models import OsfStorageFile
@@ -8,16 +9,16 @@
 from osf.metrics.utils import YearMonth
 from ._base import MonthlyReporter
 
-
 class InstitutionalSummaryMonthlyReporter(MonthlyReporter):
     """Generate an InstitutionMonthlySummaryReport for each institution."""
 
     def report(self, yearmonth: YearMonth):
         for institution in Institution.objects.all():
+            close_old_connections()
             yield self.generate_report(institution, yearmonth)
 
     def generate_report(self, institution, yearmonth):
-        node_queryset = institution.nodes.filter(
+        node_queryset = institution.nodes.select_related('type').filter(
             deleted__isnull=True,
             created__lt=yearmonth.next_month()
         ).exclude(
@@ -44,15 +45,13 @@ def _get_count(self, node_queryset, node_type, is_public):
         return node_queryset.filter(type=node_type, is_public=is_public, root_id=F('pk')).count()
 
     def get_published_preprints(self, institution, yearmonth):
-        queryset = Preprint.objects.can_view().filter(
+        return Preprint.objects.can_view().select_related('affiliated_institutions').filter(
             affiliated_institutions=institution,
             created__lte=yearmonth.next_month()
         ).exclude(
             spam_status=SpamStatus.SPAM
         )
 
-        return queryset
-
     def get_files(self, node_queryset, preprint_queryset, is_public=None):
         public_kwargs = {}
         if is_public: