From 2d63a0eba198a94ac0d1366b578b0c9055a95443 Mon Sep 17 00:00:00 2001
From: Dominic Robinson <65237317+drobinson-moj@users.noreply.github.com>
Date: Wed, 4 Dec 2024 13:14:14 +0000
Subject: [PATCH] TM-769: record endpoint monitoring response time (#1152)

---
 .../templates/collectd_endpoint_monitoring.sh.j2  | 15 +++++++++++++--
 1 file changed, 13 insertions(+), 2 deletions(-)

diff --git a/ansible/roles/collectd-endpoint-monitoring/templates/collectd_endpoint_monitoring.sh.j2 b/ansible/roles/collectd-endpoint-monitoring/templates/collectd_endpoint_monitoring.sh.j2
index 556579dcf..445255591 100644
--- a/ansible/roles/collectd-endpoint-monitoring/templates/collectd_endpoint_monitoring.sh.j2
+++ b/ansible/roles/collectd-endpoint-monitoring/templates/collectd_endpoint_monitoring.sh.j2
@@ -60,11 +60,15 @@ check_endpoint() {
   if [[ $follow_redirect == 1 ]]; then
     optional_curl_args="-L"
   fi
-  if ! output=$(curl -sSv -m "$timeout_secs" -o /dev/null -w "http_code=%{http_code}" $optional_curl_args "$url" 2>&1); then
-    grep -v "^\*" <<<"$output" | grep -v ^http_code= | grep -v CApath | grep -v "^>" | grep -v "^<" | grep -v "^{" | grep -v "^}" >&2
+  if ! output=$(curl -sSv -m "$timeout_secs" -o /dev/null -w "http_code=%{http_code}\nresponse_time=%{time_total}\n" $optional_curl_args "$url" 2>&1); then
+    grep -v "^\*" <<<"$output" | grep -v ^http_code= | grep -v ^response_time= | grep -v CApath | grep -v "^>" | grep -v "^<" | grep -v "^{" | grep -v "^}" >&2
     return 1
   fi
+  response_time=$(grep "^response_time=" <<< "$output" | cut -d= -f2)
   http_code=$(grep "^http_code=" <<< "$output" | cut -d= -f2)
+  if [[ -n $response_time ]]; then
+    echo "response_time=$response_time"
+  fi
   if [[ -z $http_code ]]; then
     echo "missing http_code in curl output" >&2
     return 1
@@ -135,12 +139,19 @@ while true; do
       exitcode=$?
     fi
     days_to_expiry=$(grep "^days_to_expiry=" <<< "$output" | cut -d= -f2)
+    response_time=$(grep "^response_time=" <<< "$output" | cut -d= -f2)
     echo "PUTVAL $HOSTNAME/endpoint_status/exitcode-${args[3]} interval=$INTERVAL N:$exitcode"
     if [[ -n $days_to_expiry ]]; then
       if [[ $log_cert_expiry -eq 1 ]]; then
         echo "PUTVAL $HOSTNAME/endpoint_cert_expiry/gauge-${args[3]} interval=$INTERVAL N:$days_to_expiry"
       fi
     fi
+    if [[ -n $response_time ]]; then
+      response_time_ms=$(echo "$response_time*1000" | bc | cut -d. -f1)
+      if [[ -n $response_time_ms ]]; then
+        echo "PUTVAL $HOSTNAME/endpoint_response_time_ms/gauge-${args[3]} interval=$INTERVAL N:$response_time_ms"
+      fi
+    fi
     if [[ $exitcode -ne 0 ]]; then
       if [[ $LOGGER_INTERVAL_FOR_ERRORS -eq 0 ]]; then
         echo "${args[3]}: $output"