chore: Modernize discourse-mixin

.gitignore

@@ -2,3 +2,4 @@
 vendor
 jsonnetfile.lock.json
 *.zip
+.worktrees

discourse-mixin/.lint

@@ -10,3 +10,5 @@ exclusions:
       - panel: "Sidekiq Workers"
   template-instance-rule:
     reason: "Based on new convention we are using variable names prometheus_datasource and loki_datasource where as linter expects 'datasource'"
+  panel-datasource-rule:
+    reason: "Modern mixins use signal-based architecture where datasource references are handled by the framework"

discourse-mixin/alerts/alerts.libsonnet → discourse-mixin/alerts.libsonnet

@@ -1,41 +1,41 @@
 {
-  prometheusAlerts+:: {
-    groups+: [
+  new(this): {
+    groups: [
       {
-        name: 'DiscourseAlerts',
+        name: this.config.uid + '-alerts',
         rules: [
           {
-            alert: 'DiscourseRequestsHigh5xxErrors',
+            alert: 'DiscourseHigh5xxErrors',
             expr: |||
               100 * rate(discourse_http_requests{status="500"}[5m]) / on() group_left() (sum(rate(discourse_http_requests[5m])) by (instance)) > %(alertsCritical5xxResponses)s
-            ||| % $._config,
+            ||| % this.config,
             'for': '5m',
             labels: {
               severity: 'critical',
             },
             annotations: {
-              summary: 'More than %(alertsCritical5xxResponses)s%% of all requests result in a 5XX.' % $._config,
+              summary: 'More than %(alertsCritical5xxResponses)s%% of all requests result in a 5XX.' % this.config,
               description:
                 ('{{ printf "%%.2f" $value }}%% of all requests are resulting in 500 status codes, ' +
                  'which is above the threshold %(alertsCritical5xxResponses)s%%, ' +
-                 'indicating a potentially larger issue for {{$labels.instance}}') % $._config,
+                 'indicating a potentially larger issue for {{$labels.instance}}') % this.config,
             },
           },
           {
-            alert: 'DiscourseRequestsHigh4xxErrors',
+            alert: 'DiscourseHigh4xxErrors',
             expr: |||
               100 * rate(discourse_http_requests{status=~"^4.*"}[5m]) / on() group_left() (sum(rate(discourse_http_requests[5m])) by (instance)) > %(alertsWarning4xxResponses)s
-            ||| % $._config,
+            ||| % this.config,
             'for': '5m',
             labels: {
               severity: 'warning',
             },
             annotations: {
-              summary: 'More than %(alertsWarning4xxResponses)s%% of all requests result in a 4XX.' % $._config,
+              summary: 'More than %(alertsWarning4xxResponses)s%% of all requests result in a 4XX.' % this.config,
               description:
                 ('{{ printf "%%.2f" $value }}%% of all requests are resulting in 400 status code, ' +
                  'which is above the threshold %(alertsWarning4xxResponses)s%%, ' +
-                 'indicating a potentially larger issue for {{$labels.instance}}') % $._config,
+                 'indicating a potentially larger issue for {{$labels.instance}}') % this.config,
             },
           },
         ],

discourse-mixin/config.libsonnet

@@ -1,12 +1,34 @@
 {
-  _config+:: {
-    dashboardTags: ['discourse-mixin'],
-    dashboardPeriod: 'now-1h',
-    dashboardTimezone: 'default',
-    dashboardRefresh: '1m',
-
-    // for alerts
-    alertsCritical5xxResponses: '10',  // %
-    alertsWarning4xxResponses: '30',  // %
+  local this = self,
+
+  // Filtering
+  filteringSelector: 'job=~"$job", instance=~"$instance"',
+  groupLabels: ['job'],
+  instanceLabels: ['instance'],
+
+  // Dashboard settings
+  dashboardTags: ['discourse-mixin'],
+  dashboardPeriod: 'now-1h',
+  dashboardTimezone: 'default',
+  dashboardRefresh: '1m',
+  dashboardNamePrefix: 'Discourse',
+  uid: 'discourse',
+
+  // Logs configuration
+  enableLokiLogs: false,
+
+  // Alert thresholds
+  alertsCritical5xxResponses: 10,  // %
+  alertsWarning4xxResponses: 30,  // %
+
+  // Metrics source
+  metricsSource: 'prometheus',
+
+  // Signal categories
+  signals: {
+    http: (import './signals/http.libsonnet')(this),
+    requests: (import './signals/requests.libsonnet')(this),
+    jobs: (import './signals/jobs.libsonnet')(this),
+    memory: (import './signals/memory.libsonnet')(this),
   },
 }

discourse-mixin/dashboards.libsonnet

@@ -0,0 +1,86 @@
+local g = import './g.libsonnet';
+local commonlib = import 'common-lib/common/main.libsonnet';
+
+{
+  local root = self,
+  new(this):
+    local prefix = this.config.dashboardNamePrefix;
+    local links = this.grafana.links;
+    local tags = this.config.dashboardTags;
+    local uid = this.config.uid;
+    local vars = commonlib.variables.new(
+      filteringSelector=this.config.filteringSelector,
+      groupLabels=this.config.groupLabels,
+      instanceLabels=this.config.instanceLabels,
+      varMetric='discourse_page_views',
+      customAllValue='.+',
+      enableLokiLogs=this.config.enableLokiLogs,
+    );
+    local annotations = {};
+    local refresh = this.config.dashboardRefresh;
+    local period = this.config.dashboardPeriod;
+    local timezone = this.config.dashboardTimezone;
+
+    {
+      'discourse-overview.json':
+        g.dashboard.new(prefix + ' overview')
+        + g.dashboard.withDescription('Overview of Discourse application performance and traffic.')
+        + g.dashboard.withPanels(
+          g.util.panel.resolveCollapsedFlagOnRows(
+            g.util.grid.wrapPanels(
+              [
+                this.grafana.rows.overviewRow,
+                this.grafana.rows.latencyRow,
+              ]
+            )
+          )
+        )
+        + root.applyCommon(
+          vars.multiInstance,
+          uid + '-overview',
+          tags,
+          links { overview+:: {} },
+          annotations,
+          timezone,
+          refresh,
+          period
+        ),
+
+      'discourse-jobs.json':
+        g.dashboard.new(prefix + ' jobs processing')
+        + g.dashboard.withDescription('Discourse job processing, workers, and memory usage.')
+        + g.dashboard.withPanels(
+          g.util.panel.resolveCollapsedFlagOnRows(
+            g.util.grid.wrapPanels(
+              [
+                this.grafana.rows.jobStatsRow,
+                this.grafana.rows.jobCountsRow,
+                this.grafana.rows.jobDurationRow,
+                this.grafana.rows.memoryRow,
+              ]
+            )
+          )
+        )
+        + root.applyCommon(
+          vars.multiInstance,
+          uid + '-jobs',
+          tags,
+          links { jobs+:: {} },
+          annotations,
+          timezone,
+          refresh,
+          period
+        ),
+    },
+
+  applyCommon(vars, uid, tags, links, annotations, timezone, refresh, period):
+    g.dashboard.withTags(tags)
+    + g.dashboard.withUid(uid)
+    + g.dashboard.withLinks(std.objectValues(links))
+    + g.dashboard.withTimezone(timezone)
+    + g.dashboard.withRefresh(refresh)
+    + g.dashboard.time.withFrom(period)
+    + g.dashboard.withVariables(vars)
+    + g.dashboard.withAnnotations(std.objectValues(annotations))
+    + g.dashboard.graphTooltip.withSharedCrosshair(),
+}