fix(gcp): correct NRQL aggregators, scaling, and gcpSample alignment across 36 entity types

entity-types/infra-gcpaiplatformendpoint/golden_metrics.yml

@@ -3,7 +3,7 @@ cpuUtilization:
   unit: PERCENTAGE
   queries:
     gcp:
-      select: average(gcp.aiplatform.prediction.online.cpu.utilization)
+      select: average(gcp.aiplatform.prediction.online.cpu.utilization) * 100
       from: Metric
       eventId: entity.guid
       eventName: entity.name
@@ -31,12 +31,12 @@ bytesSent:
   unit: BYTES
   queries:
     gcp:
-      select: average(gcp.aiplatform.prediction.online.network.sent_bytes_count)
+      select: sum(gcp.aiplatform.prediction.online.network.sent_bytes_count)
       from: Metric
       eventId: entity.guid
       eventName: entity.name
     gcpSample:
-      select: average(prediction.online.network.SentBytes)
+      select: sum(prediction.online.network.SentBytes)
       from: GcpVertexAiEndpointSample
       eventId: entityGuid
       eventName: entityName
@@ -45,12 +45,12 @@ bytesReceived:
   unit: BYTES
   queries:
     gcp:
-      select: average(gcp.aiplatform.prediction.online.network.received_bytes_count)
+      select: sum(gcp.aiplatform.prediction.online.network.received_bytes_count)
       from: Metric
       eventId: entity.guid
       eventName: entity.name
     gcpSample:
-      select: average(prediction.online.network.ReceivedBytes)
+      select: sum(prediction.online.network.ReceivedBytes)
       from: GcpVertexAiEndpointSample
       eventId: entityGuid
       eventName: entityName

entity-types/infra-gcpaiplatformfeatureonlinestore/golden_metrics.yml

@@ -13,16 +13,16 @@ request:
       eventId: entity.guid
       eventName: entity.name
 latencies:
-  title: Latency
+  title: Latency p95 (ms)
   unit: MS
   queries:
     gcpSample:
-      select: average(featureonlinestore.onlineServing.ServingLatencies)
+      select: percentile(featureonlinestore.onlineServing.ServingLatencies, 95)
       from: GcpVertexAiFeatureOnlineStoreSample
       eventId: entityGuid
       eventName: entityName
     gcp:
-      select: average(gcp.aiplatform.featureonlinestore.online_serving.serving_latencies)
+      select: percentile(gcp.aiplatform.featureonlinestore.online_serving.serving_latencies, 95)
       from: Metric
       eventId: entity.guid
       eventName: entity.name
@@ -54,7 +54,7 @@ bigtableCpuLoad:
   unit: PERCENTAGE
   queries:
     gcp:
-      select: average(gcp.aiplatform.featureonlinestore.storage.bigtable_cpu_load)
+      select: average(gcp.aiplatform.featureonlinestore.storage.bigtable_cpu_load) * 100
       from: Metric
       eventId: entity.guid
       eventName: entity.name
@@ -63,7 +63,7 @@ bigtableNodes:
   unit: COUNT
   queries:
     gcp:
-      select: sum(gcp.aiplatform.featureonlinestore.storage.bigtable_nodes)
+      select: average(gcp.aiplatform.featureonlinestore.storage.bigtable_nodes)
       from: Metric
       eventId: entity.guid
       eventName: entity.name

entity-types/infra-gcpaiplatformfeaturestore/golden_metrics.yml

@@ -17,7 +17,7 @@ cpuLoad:
   unit: PERCENTAGE
   queries:
     gcp:
-      select: average(gcp.aiplatform.featurestore.cpu_load)
+      select: average(gcp.aiplatform.featurestore.cpu_load) * 100
       from: Metric
       eventId: entity.guid
       eventName: entity.name
@@ -27,16 +27,16 @@ cpuLoad:
       eventId: entityGuid
       eventName: entityName
 latency:
-  title: Latency
+  title: Latency p95 (ms)
   unit: MS
   queries:
     gcp:
-      select: average(gcp.aiplatform.featurestore.online_serving.latencies)
+      select: percentile(gcp.aiplatform.featurestore.online_serving.latencies, 95)
       from: Metric
       eventId: entity.guid
       eventName: entity.name
     gcpSample:
-      select: average(featurestore.online_serving.Latencies)
+      select: percentile(featurestore.online_serving.Latencies, 95)
       from: GcpVertexAiFeaturestoreSample
       eventId: entityGuid
       eventName: entityName

entity-types/infra-gcpaiplatformindex/golden_metrics.yml

@@ -13,16 +13,16 @@ request:
       eventId: entityGuid
       eventName: entityName
 latencies:
-  title: Latency
+  title: Latency p95 (ms)
   unit: MS
   queries:
     gcp:
-      select: average(gcp.aiplatform.matching_engine.stream_update.latencies)
+      select: percentile(gcp.aiplatform.matching_engine.stream_update.latencies, 95)
       from: Metric
       eventId: entity.guid
       eventName: entity.name
     gcpSample:
-      select: average(matchingEngine.streamUpdate.Latencies)
+      select: percentile(matchingEngine.streamUpdate.Latencies, 95)
       from: GcpVertexAiIndexSample
       eventId: entityGuid
       eventName: entityName
@@ -31,12 +31,12 @@ datapointCount:
   unit: COUNT
   queries:
     gcp:
-      select: average(gcp.aiplatform.matching_engine.stream_update.datapoint_count)
+      select: sum(gcp.aiplatform.matching_engine.stream_update.datapoint_count)
       from: Metric
       eventId: entity.guid
       eventName: entity.name
     gcpSample:
-      select: average(matchingEngine.streamUpdate.Datapoint)
+      select: sum(matchingEngine.streamUpdate.Datapoint)
       from: GcpVertexAiIndexSample
       eventId: entityGuid
       eventName: entityName

entity-types/infra-gcpaiplatformlocation/golden_metrics.yml

@@ -3,7 +3,7 @@ executingPipelineJobs:
   unit: COUNT
   queries:
     gcp:
-      select: sum(gcp.aiplatform.executing_vertexai_pipeline_jobs)
+      select: average(gcp.aiplatform.executing_vertexai_pipeline_jobs)
       from: Metric
       eventId: entity.guid
       eventName: entity.name
@@ -13,12 +13,12 @@ pipelineTasks:
   unit: COUNT
   queries:
     gcpSample:
-      select: sum(executingVertexaiPipelineTasks)
+      select: average(executingVertexaiPipelineTasks)
       from: GcpVertexAiLocationSample
       eventId: entityGuid
       eventName: entityName
     gcp:
-      select: sum(gcp.aiplatform.executing_vertexai_pipeline_tasks)
+      select: average(gcp.aiplatform.executing_vertexai_pipeline_tasks)
       from: Metric
       eventId: entity.guid
       eventName: entity.name
@@ -28,12 +28,12 @@ onlinePredictionRequestsPerBaseModel:
   unit: COUNT
   queries:
     gcpSample:
-      select: average(onlinePredictionRequestsPerBaseModel)
+      select: sum(onlinePredictionRequestsPerBaseModel)
       from: GcpVertexAiLocationSample
       eventId: entityGuid
       eventName: entityName
     gcp:
-      select: average(gcp.aiplatform.online_prediction_requests_per_base_model)
+      select: sum(gcp.aiplatform.online_prediction_requests_per_base_model)
       from: Metric
       eventId: entity.guid
       eventName: entity.name

entity-types/infra-gcpalloydbcluster/golden_metrics.yml

@@ -3,7 +3,7 @@ storageUsage:
   unit: BYTES
   queries:
     gcp:
-      select: sum(gcp.alloydb.cluster.storage.usage)
+      select: latest(gcp.alloydb.cluster.storage.usage)
       from: Metric
       eventId: entity.guid
       eventName: entity.name
@@ -12,7 +12,7 @@ backupLastBackupTime:
   unit: TIMESTAMP
   queries:
     gcp:
-      select: latest(gcp.alloydb.cluster.last_backup_timestamp)
+      select: latest(gcp.alloydb.cluster.last_backup_timestamp)/1000
       from: Metric
       eventId: entity.guid
       eventName: entity.name

entity-types/infra-gcpalloydbinstance/golden_metrics.yml

@@ -3,7 +3,7 @@ cpuAverageUtilization:
   unit: PERCENTAGE
   queries:
     gcp:
-      select: average(`gcp.alloydb.instance.cpu.average_utilization`)
+      select: average(`gcp.alloydb.instance.cpu.average_utilization`) * 100
       from: Metric
       eventId: entity.guid
       eventName: entity.name
@@ -12,7 +12,7 @@ totalConnections:
   unit: COUNT
   queries:
     gcp:
-      select: sum(`gcp.alloydb.instance.postgres.total_connections`)
+      select: average(`gcp.alloydb.instance.postgres.total_connections`)
       from: Metric
       eventId: entity.guid
       eventName: entity.name

entity-types/infra-gcpalloydbinstancenode/golden_metrics.yml

@@ -3,7 +3,7 @@ cpuUsageTime:
   unit: PERCENTAGE
   queries:
     gcp:
-      select: average(`gcp.alloydb.node.cpu.usage_time`)
+      select: average(`gcp.alloydb.node.cpu.usage_time`) * 100
       from: Metric
       eventId: entity.guid
       eventName: entity.name
@@ -12,7 +12,7 @@ backends:
   unit: COUNT
   queries:
     gcp:
-      select: sum(`gcp.alloydb.node.postgres.backends`)
+      select: average(`gcp.alloydb.node.postgres.backends`)
       from: Metric
       eventId: entity.guid
       eventName: entity.name

entity-types/infra-gcpappengineservice/golden_metrics.yml

@@ -41,16 +41,16 @@ ErrorRate:
       eventId: entityGuid
       eventName: entityName
 latency:
-  title: Latency
+  title: Latency p95 (s)
   unit: SECONDS
   queries:
     gcp:
-      select: average(gcp.appengine.http.server.response_latencies) / 1000
+      select: percentile(gcp.appengine.http.server.response_latencies, 95) / 1000
       from: Metric
       eventId: entity.guid
       eventName: entity.name
     gcpSample:
-      select: average(`server.ResponseLatenciesMilliseconds`) / 1000
+      select: percentile(`server.ResponseLatenciesMilliseconds`, 95) / 1000
       from: GcpAppEngineServiceSample
       eventId: entityGuid
       eventName: entityName

entity-types/infra-gcpbigquerydataset/golden_metrics.yml

@@ -3,12 +3,12 @@ storedBytes:
   unit: BYTES
   queries:
     gcp:
-      select: sum(gcp.bigquery.storage.stored_bytes)
+      select: latest(gcp.bigquery.storage.stored_bytes)
       from: Metric
       eventId: entity.guid
       eventName: entity.name
     gcpSample:
-      select: sum(`storage.StoredBytes`)
+      select: latest(`storage.StoredBytes`)
       from: GcpBigQueryDataSetSample
       eventId: entityGuid
       eventName: entityName
@@ -17,12 +17,12 @@ tables:
   unit: COUNT
   queries:
     gcp:
-      select: sum(gcp.bigquery.storage.table_count)
+      select: average(gcp.bigquery.storage.table_count)
       from: Metric
       eventId: entity.guid
       eventName: entity.name
     gcpSample:
-      select: sum(`storage.Tables`)
+      select: average(`storage.Tables`)
       from: GcpBigQueryDataSetSample
       eventId: entityGuid
       eventName: entityName

entity-types/infra-gcpbigqueryproject/golden_metrics.yml

@@ -3,26 +3,26 @@ queries:
   unit: COUNT
   queries:
     gcp:
-      select: sum(`gcp.bigquery.query.count`)
+      select: average(`gcp.bigquery.query.count`)
       from: Metric
       eventId: entity.guid
       eventName: entity.name
     gcpSample:
-      select: sum(`query.Count`)
+      select: average(`query.Count`)
       from: GcpBigQueryProjectSample
       eventId: entityGuid
       eventName: entityName
 executionTime:
-  title: Execution time
+  title: Execution time p95 (s)
   unit: SECONDS
   queries:
     gcp:
-      select: average(`gcp.bigquery.query.execution_times`)
+      select: percentile(`gcp.bigquery.query.execution_times`, 95)
       from: Metric
       eventId: entity.guid
       eventName: entity.name
     gcpSample:
-      select: average(`query.ExecutionTimes`)
+      select: percentile(`query.ExecutionTimes`, 95)
       from: GcpBigQueryProjectSample
       eventId: entityGuid
       eventName: entityName

entity-types/infra-gcpbigtablecluster/golden_metrics.yml

@@ -3,7 +3,7 @@ cpuLoad:
   unit: PERCENTAGE
   queries:
     gcp:
-      select: average(gcp.bigtable.cluster.cpuLoad)
+      select: average(gcp.bigtable.cluster.cpu_load) * 100
       from: Metric
       eventId: entity.guid
       eventName: entity.name
@@ -17,7 +17,7 @@ cpuLoadOfBusiestNode:
   unit: PERCENTAGE
   queries:
     gcp:
-      select: average(gcp.bigtable.cluster.cpuLoadHottestNode)
+      select: average(gcp.bigtable.cluster.cpu_load_hottest_node) * 100
       from: Metric
       eventId: entity.guid
       eventName: entity.name
@@ -31,7 +31,7 @@ utilizationOfHddDisks:
   unit: PERCENTAGE
   queries:
     gcp:
-      select: average(gcp.bigtable.cluster.diskLoad)
+      select: average(gcp.bigtable.cluster.disk_load) * 100
       from: Metric
       eventId: entity.guid
       eventName: entity.name
@@ -45,7 +45,7 @@ storageUsed:
   unit: PERCENTAGE
   queries:
     gcp:
-      select: average(gcp.bigtable.cluster.storageUtilization)
+      select: average(gcp.bigtable.cluster.storage_utilization) * 100
       from: Metric
       eventId: entity.guid
       eventName: entity.name