From 237264a6a87cf6f44a1e5a34e595a82d7be0f718 Mon Sep 17 00:00:00 2001
From: rsora <r.sora@arduino.cc>
Date: Fri, 24 Sep 2021 15:44:15 +0200
Subject: [PATCH] [skip changelog] Add partition load query before fetching
 metrics and schedule extraction 3 times a day

---
 .github/tools/fetch_athena_stats.sh  | 29 +++++++++++++++++++++++++++-
 .github/workflows/arduino-stats.yaml |  4 ++--
 2 files changed, 30 insertions(+), 3 deletions(-)

diff --git a/.github/tools/fetch_athena_stats.sh b/.github/tools/fetch_athena_stats.sh
index b7010a8235b..e04ca51d38a 100755
--- a/.github/tools/fetch_athena_stats.sh
+++ b/.github/tools/fetch_athena_stats.sh
@@ -14,6 +14,34 @@
 
 set -euo pipefail
 
+loadExecutionId=$(
+  aws athena start-query-execution \
+    --query-string "MSCK REPAIR TABLE ${AWS_ATHENA_SOURCE_TABLE};" \
+    --result-configuration "OutputLocation=${AWS_ATHENA_OUTPUT_LOCATION}" \
+    --region us-east-1 | jq -r ".QueryExecutionId"
+)
+
+echo "QueryExecutionId is ${loadExecutionId}"
+for i in $(seq 1 120); do
+  loadState=$(
+    aws athena get-query-execution \
+      --query-execution-id "${loadExecutionId}" \
+      --region us-east-1 | jq -r ".QueryExecution.Status.State"
+  )
+
+  if [[ "${loadState}" == "SUCCEEDED" ]]; then
+    break
+  fi
+
+  echo "QueryExecutionId ${loadExecutionId} - state is ${loadState}"
+
+  if [[ "${loadState}" == "FAILED" ]]; then
+    exit 1
+  fi
+
+  sleep 2
+done
+
 ! read -r -d '' query <<EOM
 SELECT replace(json_extract_scalar(url_decode(url_decode(querystring)),
         '$.data.url'), 'https://downloads.arduino.cc/arduino-cli/arduino-cli_', '') AS flavor, count(json_extract(url_decode(url_decode(querystring)),'$')) AS gauge
@@ -28,7 +56,6 @@ EOM
 queryExecutionId=$(
   aws athena start-query-execution \
     --query-string "${query}" \
-    --query-execution-context "Database=demo_books" \
     --result-configuration "OutputLocation=${AWS_ATHENA_OUTPUT_LOCATION}" \
     --region us-east-1 | jq -r ".QueryExecutionId"
 )
diff --git a/.github/workflows/arduino-stats.yaml b/.github/workflows/arduino-stats.yaml
index 73bd7ff3abd..b8799f3bea3 100644
--- a/.github/workflows/arduino-stats.yaml
+++ b/.github/workflows/arduino-stats.yaml
@@ -2,8 +2,8 @@ name: arduino-stats
 
 on:
   schedule:
-    # run every day at 12:30:00
-    - cron: "30 12 * * *"
+    # run every day at 07:00 AM, 03:00 PM and 11:00 PM
+    - cron: "0 7,15,23 * * *"
   workflow_dispatch:
   repository_dispatch: