diff --git a/operations/helm/charts/mimir-distributed/templates/alertmanager/alertmanager-statefulset.yaml b/operations/helm/charts/mimir-distributed/templates/alertmanager/alertmanager-statefulset.yaml
index 5437fc483db..4e157d904b8 100644
--- a/operations/helm/charts/mimir-distributed/templates/alertmanager/alertmanager-statefulset.yaml
+++ b/operations/helm/charts/mimir-distributed/templates/alertmanager/alertmanager-statefulset.yaml
@@ -147,11 +147,11 @@ spec:
             {{- if .Values.alertmanager.zoneAwareReplication.enabled }}
             - "-alertmanager.sharding-ring.instance-availability-zone=zone-default"
             {{- end }}
+            {{- end }}
             # Prometheus HTTP client used to send alerts has a hard-coded idle
             # timeout of 5 minutes, therefore the server timeout for Alertmanager
             # needs to be higher to avoid connections being closed abruptly.
             - "-server.http-idle-timeout=6m"
-            {{- end }}
             {{- range $key, $value := .Values.alertmanager.extraArgs }}
             - "-{{ $key }}={{ $value }}"
             {{- end }}
diff --git a/operations/helm/tests/test-oss-logical-multizone-values-generated/mimir-distributed/templates/alertmanager/alertmanager-statefulset.yaml b/operations/helm/tests/test-oss-logical-multizone-values-generated/mimir-distributed/templates/alertmanager/alertmanager-statefulset.yaml
index 81d87150eab..1997b823720 100644
--- a/operations/helm/tests/test-oss-logical-multizone-values-generated/mimir-distributed/templates/alertmanager/alertmanager-statefulset.yaml
+++ b/operations/helm/tests/test-oss-logical-multizone-values-generated/mimir-distributed/templates/alertmanager/alertmanager-statefulset.yaml
@@ -87,6 +87,10 @@ spec:
             - "-config.expand-env=true"
             - "-config.file=/etc/mimir/mimir.yaml"
             - "-alertmanager.sharding-ring.instance-availability-zone=zone-a"
+            # Prometheus HTTP client used to send alerts has a hard-coded idle
+            # timeout of 5 minutes, therefore the server timeout for Alertmanager
+            # needs to be higher to avoid connections being closed abruptly.
+            - "-server.http-idle-timeout=6m"
           volumeMounts:
             - name: config
               mountPath: /etc/mimir
@@ -214,6 +218,10 @@ spec:
             - "-config.expand-env=true"
             - "-config.file=/etc/mimir/mimir.yaml"
             - "-alertmanager.sharding-ring.instance-availability-zone=zone-b"
+            # Prometheus HTTP client used to send alerts has a hard-coded idle
+            # timeout of 5 minutes, therefore the server timeout for Alertmanager
+            # needs to be higher to avoid connections being closed abruptly.
+            - "-server.http-idle-timeout=6m"
           volumeMounts:
             - name: config
               mountPath: /etc/mimir
@@ -341,6 +349,10 @@ spec:
             - "-config.expand-env=true"
             - "-config.file=/etc/mimir/mimir.yaml"
             - "-alertmanager.sharding-ring.instance-availability-zone=zone-c"
+            # Prometheus HTTP client used to send alerts has a hard-coded idle
+            # timeout of 5 minutes, therefore the server timeout for Alertmanager
+            # needs to be higher to avoid connections being closed abruptly.
+            - "-server.http-idle-timeout=6m"
           volumeMounts:
             - name: config
               mountPath: /etc/mimir
diff --git a/operations/helm/tests/test-oss-multizone-values-generated/mimir-distributed/templates/alertmanager/alertmanager-statefulset.yaml b/operations/helm/tests/test-oss-multizone-values-generated/mimir-distributed/templates/alertmanager/alertmanager-statefulset.yaml
index 9f6d1116ed7..d94716ea17d 100644
--- a/operations/helm/tests/test-oss-multizone-values-generated/mimir-distributed/templates/alertmanager/alertmanager-statefulset.yaml
+++ b/operations/helm/tests/test-oss-multizone-values-generated/mimir-distributed/templates/alertmanager/alertmanager-statefulset.yaml
@@ -110,6 +110,10 @@ spec:
             - "-config.expand-env=true"
             - "-config.file=/etc/mimir/mimir.yaml"
             - "-alertmanager.sharding-ring.instance-availability-zone=zone-a"
+            # Prometheus HTTP client used to send alerts has a hard-coded idle
+            # timeout of 5 minutes, therefore the server timeout for Alertmanager
+            # needs to be higher to avoid connections being closed abruptly.
+            - "-server.http-idle-timeout=6m"
           volumeMounts:
             - name: config
               mountPath: /etc/mimir
@@ -262,6 +266,10 @@ spec:
             - "-config.expand-env=true"
             - "-config.file=/etc/mimir/mimir.yaml"
             - "-alertmanager.sharding-ring.instance-availability-zone=zone-b"
+            # Prometheus HTTP client used to send alerts has a hard-coded idle
+            # timeout of 5 minutes, therefore the server timeout for Alertmanager
+            # needs to be higher to avoid connections being closed abruptly.
+            - "-server.http-idle-timeout=6m"
           volumeMounts:
             - name: config
               mountPath: /etc/mimir
@@ -414,6 +422,10 @@ spec:
             - "-config.expand-env=true"
             - "-config.file=/etc/mimir/mimir.yaml"
             - "-alertmanager.sharding-ring.instance-availability-zone=zone-c"
+            # Prometheus HTTP client used to send alerts has a hard-coded idle
+            # timeout of 5 minutes, therefore the server timeout for Alertmanager
+            # needs to be higher to avoid connections being closed abruptly.
+            - "-server.http-idle-timeout=6m"
           volumeMounts:
             - name: config
               mountPath: /etc/mimir