Merge pull request #3 from observIQ/fix-load-metric-units

Fix warehouse load metric unit descriptions
grafana · Jan 11, 2023 · a8e2831 · a8e2831
2 parents 8993dd7 + 8aa49dc
commit a8e2831
Show file tree

Hide file tree

Showing 3 changed files with 36 additions and 25 deletions.
diff --git a/collector/collector.go b/collector/collector.go
@@ -66,10 +66,10 @@ type Collector struct {
 	logins                            *prometheus.Desc
 	successfulLogins                  *prometheus.Desc
 	failedLogins                      *prometheus.Desc
-	warehouseExecutedQueries          *prometheus.Desc
-	warehouseOverloadedQueueSize      *prometheus.Desc
-	warehouseProvisioningQueueSize    *prometheus.Desc
-	warehouseBlockedQueries           *prometheus.Desc
+	warehouseExecutedQueryLoad        *prometheus.Desc
+	warehouseOverloadedQueueLoad      *prometheus.Desc
+	warehouseProvisioningQueueLoad    *prometheus.Desc
+	warehouseBlockedQueryLoad         *prometheus.Desc
 	autoClusteringCredits             *prometheus.Desc
 	autoClusteringBytes               *prometheus.Desc
 	autoClusteringRows                *prometheus.Desc
@@ -161,27 +161,27 @@ func NewCollector(logger log.Logger, c *Config) *Collector {
 			[]string{labelClientType, labelClientVersion},
 			nil,
 		),
-		warehouseExecutedQueries: prometheus.NewDesc(
+		warehouseExecutedQueryLoad: prometheus.NewDesc(
 			prometheus.BuildFQName(namespace, "warehouse", "executed_queries"),
-			"Average number of queries executed.",
+			"Average query load for queries executed over the last 24 hours.",
 			[]string{labelName, labelID},
 			nil,
 		),
-		warehouseOverloadedQueueSize: prometheus.NewDesc(
+		warehouseOverloadedQueueLoad: prometheus.NewDesc(
 			prometheus.BuildFQName(namespace, "warehouse", "overloaded_queue_size"),
-			"Average number of queries queued because the warehouse was overloaded.",
+			"Average load value for queries queued because the warehouse was being overloaded over the last 24 hours.",
 			[]string{labelName, labelID},
 			nil,
 		),
-		warehouseProvisioningQueueSize: prometheus.NewDesc(
+		warehouseProvisioningQueueLoad: prometheus.NewDesc(
 			prometheus.BuildFQName(namespace, "warehouse", "provisioning_queue_size"),
-			"Average number of queries queued because the warehouse was being provisioned.",
+			"Average load value for queries queued because the warehouse was being provisioned over the last 24 hours.",
 			[]string{labelName, labelID},
 			nil,
 		),
-		warehouseBlockedQueries: prometheus.NewDesc(
+		warehouseBlockedQueryLoad: prometheus.NewDesc(
 			prometheus.BuildFQName(namespace, "warehouse", "blocked_queries"),
-			"Average number of queries blocked by a transaction lock.",
+			"Average load value for queries blocked by a transaction lock over the last 24 hours.",
 			[]string{labelName, labelID},
 			nil,
 		),
@@ -264,10 +264,10 @@ func (c *Collector) Describe(descs chan<- *prometheus.Desc) {
 	descs <- c.logins
 	descs <- c.successfulLogins
 	descs <- c.failedLogins
-	descs <- c.warehouseExecutedQueries
-	descs <- c.warehouseOverloadedQueueSize
-	descs <- c.warehouseProvisioningQueueSize
-	descs <- c.warehouseBlockedQueries
+	descs <- c.warehouseExecutedQueryLoad
+	descs <- c.warehouseOverloadedQueueLoad
+	descs <- c.warehouseProvisioningQueueLoad
+	descs <- c.warehouseBlockedQueryLoad
 	descs <- c.autoClusteringCredits
 	descs <- c.autoClusteringBytes
 	descs <- c.autoClusteringRows
@@ -470,10 +470,10 @@ func (c *Collector) collectWarehouseLoadMetrics(db *sql.DB, metrics chan<- prome
 			return fmt.Errorf("failed to scan row: %w", err)
 		}
 
-		metrics <- prometheus.MustNewConstMetric(c.warehouseExecutedQueries, prometheus.GaugeValue, avgRunning, warehouseName, warehouseID)
-		metrics <- prometheus.MustNewConstMetric(c.warehouseOverloadedQueueSize, prometheus.GaugeValue, avgQueued, warehouseName, warehouseID)
-		metrics <- prometheus.MustNewConstMetric(c.warehouseProvisioningQueueSize, prometheus.GaugeValue, avgQueuedProvisioning, warehouseName, warehouseID)
-		metrics <- prometheus.MustNewConstMetric(c.warehouseBlockedQueries, prometheus.GaugeValue, avgBlocked, warehouseName, warehouseID)
+		metrics <- prometheus.MustNewConstMetric(c.warehouseExecutedQueryLoad, prometheus.GaugeValue, avgRunning, warehouseName, warehouseID)
+		metrics <- prometheus.MustNewConstMetric(c.warehouseOverloadedQueueLoad, prometheus.GaugeValue, avgQueued, warehouseName, warehouseID)
+		metrics <- prometheus.MustNewConstMetric(c.warehouseProvisioningQueueLoad, prometheus.GaugeValue, avgQueuedProvisioning, warehouseName, warehouseID)
+		metrics <- prometheus.MustNewConstMetric(c.warehouseBlockedQueryLoad, prometheus.GaugeValue, avgBlocked, warehouseName, warehouseID)
 	}
 
 	return rows.Err()

diff --git a/mixin/.lint b/mixin/.lint
@@ -7,3 +7,14 @@ exclusions:
     reason: "Based on new convention we are using variable names prometheus_datasource and loki_datasource where as linter expects 'datasource'"
   template-instance-rule:
     reason: "Based on new convention we are using variable names prometheus_datasource and loki_datasource where as linter expects 'datasource'"
+  panel-units-rule:
+    reason: "Custom units are used for better user experience in some panels"
+    entries:
+    - panel: "Login attempts"    
+    - panel: "Autoclustering credits used"
+    - panel: "Top 5 service compute credits used"
+    - panel: "Top 5 service cloud service credits used"
+    - panel: "Top 5 warehouse compute credits used"
+    - panel: "Top 5 warehouse cloud services credits used"
+    - panel: "Top 5 database autoclustering credits used"
+    - panel: "Top 5 table autoclustering credits used"
diff --git a/mixin/dashboards/snowflake-overview.libsonnet b/mixin/dashboards/snowflake-overview.libsonnet
@@ -18,22 +18,22 @@ local warehouseActivityPanel = {
     prometheus.target(
       'snowflake_warehouse_executed_queries{instance=~"$instance", job=~"$job", name=~"$warehouse"}',
       datasource=promDatasource,
-      legendFormat='{{instance}} - {{name}} - Executed queries'
+      legendFormat='{{instance}} - {{name}} - Executed query load'
     ),
     prometheus.target(
       'snowflake_warehouse_overloaded_queue_size{instance=~"$instance", job=~"$job", name=~"$warehouse"}',
       datasource=promDatasource,
-      legendFormat='{{instance}} - {{name}} - Overloaded queue size'
+      legendFormat='{{instance}} - {{name}} - Overloaded queue load'
     ),
     prometheus.target(
       'snowflake_warehouse_provisioning_queue_size{instance=~"$instance", job=~"$job", name=~"$warehouse"}',
       datasource=promDatasource,
-      legendFormat='{{instance}} - {{name}} - Provisioning queue size'
+      legendFormat='{{instance}} - {{name}} - Provisioning queue load'
     ),
     prometheus.target(
       'snowflake_warehouse_blocked_queries{instance=~"$instance", job=~"$job", name=~"$warehouse"}',
       datasource=promDatasource,
-      legendFormat='{{instance}} - {{name}} - Blocked queries'
+      legendFormat='{{instance}} - {{name}} - Blocked query load'
     ),
   ],
   type: 'timeseries',
@@ -78,7 +78,7 @@ local warehouseActivityPanel = {
         },
       },
       mappings: [],
-      unit: 'queries / hr',
+      unit: 'percentunit',
     },
     overrides: [],
   },