From c9e9d5d74fcfed3789189d79879797f505ab0c0b Mon Sep 17 00:00:00 2001
From: gabemontero <gmontero@redhat.com>
Date: Thu, 25 Jul 2024 12:01:00 -0400
Subject: [PATCH] SRVKP-4532: factor k8s throttling into task panel

Also:
- fix tekton-config.yaml yamllint error
- widen controller restart query
- make osp pruner cfg consistent with konflux prod

rh-pre-commit.version: 2.3.0
rh-pre-commit.check-secrets: ENABLED
---
 .../pipeline-service-dashboard.json           |  6 +++---
 .../openshift-pipelines/tekton-config.yaml    | 19 +++++++++----------
 2 files changed, 12 insertions(+), 13 deletions(-)

diff --git a/operator/gitops/argocd/grafana/dashboards/pipeline-service-dashboard.json b/operator/gitops/argocd/grafana/dashboards/pipeline-service-dashboard.json
index 59b46431f..918b559d9 100644
--- a/operator/gitops/argocd/grafana/dashboards/pipeline-service-dashboard.json
+++ b/operator/gitops/argocd/grafana/dashboards/pipeline-service-dashboard.json
@@ -227,7 +227,7 @@
       "type": "stat"
     },
     {
-      "description": "The number of times the pipelines controller has restarted",
+      "description": "The number of times any of the pipelines controllers hav restarted",
       "fieldConfig": {
         "defaults": {
           "color": {
@@ -281,7 +281,7 @@
       "targets": [
         {
           "editorMode": "code",
-          "expr": "sum(increase(kube_pod_container_status_restarts_total{namespace=\"openshift-pipelines\", pod=~\"tekton-pipelines-controller-.*\"}[2m]))",
+          "expr": "sum(increase(kube_pod_container_status_restarts_total{namespace=\"openshift-pipelines\", pod=~\"tekton-.*\"}[2m]))",
           "legendFormat": "__auto",
           "range": true,
           "refId": "A"
@@ -409,7 +409,7 @@
       "targets": [
         {
           "editorMode": "code",
-          "expr": "sum(increase(taskrun_pod_create_not_attempted_or_pending_count[2m]))",
+          "expr": "sum(increase(taskrun_pod_create_not_attempted_or_pending_count[2m])) - sum(increase(tekton_pipelines_controller_running_taskruns_throttled_by_quota[2m])) - sum(increase(tekton_pipelines_controller_running_taskruns_throttled_by_node[2m]))",
           "legendFormat": "__auto",
           "range": true,
           "refId": "A"
diff --git a/operator/gitops/argocd/pipeline-service/openshift-pipelines/tekton-config.yaml b/operator/gitops/argocd/pipeline-service/openshift-pipelines/tekton-config.yaml
index 32429e2cc..643050aad 100644
--- a/operator/gitops/argocd/pipeline-service/openshift-pipelines/tekton-config.yaml
+++ b/operator/gitops/argocd/pipeline-service/openshift-pipelines/tekton-config.yaml
@@ -190,13 +190,12 @@ spec:
       kube-api-qps: 50
       kube-api-burst: 50
   pruner:
-      # The load on prod-rh01 is to the point now where tekton-results
-      # can fall too far behind.  Until the watcher's log storage is rewritten
-      # etc with SRVKP-4347 or if we risk adding more processing power (threads,qps,burst)
-      # to the mem leak version of the watcher, we need the OSP pruner as a backup.
-      # a bit of an adjustment, we will prune once an hour now per https://crontab.guru/every-1-hour
-      # to line up with typical timeout settings.
-      keep: 10
-      resources:
-        - pipelinerun
-      schedule: 0 * * * *
+    # The load on prod-rh01 is to the point now where tekton-results
+    # can fall too far behind.  Until the watcher's log storage is rewritten
+    # etc with SRVKP-4347 or if we risk adding more processing power (threads,qps,burst)
+    # to the mem leak version of the watcher, we need the OSP pruner as a backup.
+    disable: false
+    keep-since: 60
+    resources:
+      - pipelinerun
+    schedule: "*/10 * * * *"