From 3832ff52b4445324b58a5da123ef4e3880df1591 Mon Sep 17 00:00:00 2001
From: Dan Smith <dansmith@redhat.com>
Date: Thu, 3 Aug 2023 09:16:55 -0700
Subject: [PATCH] Add SERVICE_REPORT_INTERVAL knob

Heavily-loaded workers in CI consistently fail to complete the
service checkin task, which is configured for every ten seconds in
nova and cinder. This generates additional load on the database server
as well as consumes a threadpool worker. If we're not making the
deadline, there's really no point in having it be so high. Further,
since the workers must remain up for all the tempest tests we're
running against them, there's really no benefit to a fast-fail
detection.

This sets the report_interval to 120s for nova and cinder, and sets
service_down_time to 6x that value, which is consistent with the
default scale.

Depends-On: https://review.opendev.org/c/openstack/tempest/+/890448
Change-Id: Idd7aa1daf354256b143a3778f161cfc72b318ea5
---
 lib/cinder | 8 ++++++++
 lib/nova   | 8 ++++++++
 2 files changed, 16 insertions(+)

diff --git a/lib/cinder b/lib/cinder
index e37eff4019..f8682d5a71 100644
--- a/lib/cinder
+++ b/lib/cinder
@@ -76,6 +76,11 @@ CINDER_SERVICE_PORT_INT=${CINDER_SERVICE_PORT_INT:-18776}
 CINDER_SERVICE_PROTOCOL=${CINDER_SERVICE_PROTOCOL:-$SERVICE_PROTOCOL}
 CINDER_SERVICE_LISTEN_ADDRESS=${CINDER_SERVICE_LISTEN_ADDRESS:-$(ipv6_unquote $SERVICE_LISTEN_ADDRESS)}
 
+# We do not need to report service status every 10s for devstack-like
+# deployments. In the gate this generates extra work for the services and the
+# database which are already taxed.
+CINDER_SERVICE_REPORT_INTERVAL=120
+
 # What type of LVM device should Cinder use for LVM backend
 # Defaults to auto, which will do thin provisioning if it's a fresh
 # volume group, otherwise it will do thick. The other valid choices are
@@ -325,6 +330,9 @@ function configure_cinder {
     # details and example failures.
     iniset $CINDER_CONF DEFAULT rpc_response_timeout 120
 
+    iniset $CINDER_CONF DEFAULT report_interval $CINDER_SERVICE_REPORT_INTERVAL
+    iniset $CINDER_CONF DEFAULT service_down_time $(($CINDER_SERVICE_REPORT_INTERVAL * 6))
+
     if is_service_enabled c-vol && [[ -n "$CINDER_ENABLED_BACKENDS" ]]; then
         local enabled_backends=""
         local default_name=""
diff --git a/lib/nova b/lib/nova
index 888a2e2b25..905788f28f 100644
--- a/lib/nova
+++ b/lib/nova
@@ -75,6 +75,11 @@ NOVA_API_PASTE_INI=${NOVA_API_PASTE_INI:-$NOVA_CONF_DIR/api-paste.ini}
 # mean "use uwsgi" because we'll be always using uwsgi.
 NOVA_USE_MOD_WSGI=${NOVA_USE_MOD_WSGI:-True}
 
+# We do not need to report service status every 10s for devstack-like
+# deployments. In the gate this generates extra work for the services and the
+# database which are already taxed.
+NOVA_SERVICE_REPORT_INTERVAL=120
+
 if is_service_enabled tls-proxy; then
     NOVA_SERVICE_PROTOCOL="https"
 fi
@@ -448,6 +453,9 @@ function create_nova_conf {
 
     iniset $NOVA_CONF key_manager backend nova.keymgr.conf_key_mgr.ConfKeyManager
 
+    iniset $NOVA_CONF DEFAULT report_interval $NOVA_SERVICE_REPORT_INTERVAL
+    iniset $NOVA_CONF DEFAULT service_down_time $(($NOVA_SERVICE_REPORT_INTERVAL * 6))
+
     if is_fedora; then
         # nova defaults to /usr/local/bin, but fedora pip like to
         # install things in /usr/bin