Google Cloud metrics

Cloud Monitoring supports the metric types from Google Cloud services listed on this page.

• For a general explanation of the entries in the tables, including information about values like DELTA and GAUGE, see Metric types.

  Note: Metric types with values of type STRING can't be retrieved by using the Google Cloud Console or used in alerting policies. To retrieve string-valued data, you must use a method in the Monitoring API like timeSeries.list.

• For information about the units used in the metric lists, see the unit field in the MetricDescriptor reference.

• For information about statements of the form “Sampled every x seconds” and “After sampling, data is not visible for up to y seconds”, see Additional information: metadata.

• For a set of complete, current lists of supported metric types, see Metrics list.

• For pricing information, see Monitoring pricing.

Note: To monitor the availability of a service, create an uptime check. These checks monitor the monitoring.googleapis.com/uptime_check/check_passed metric type. Don't configure an alerting policy to track a metric type such as compute.googleapis.com/instance/uptime if your goal is to monitor the availability of a service. For more information, see Managing uptime checks.

Google Cloud metrics

Metrics from Google Cloud.

The following list was last generated at 2021-02-25 15:40:33 UTC. For more information about this process, see About the lists.

actions

Metrics from Google Assistant Smart Home. Launch stages of these metrics: BETA

The "metric type" strings in this table must be prefixed with actions.googleapis.com/. That prefix has been omitted from the entries in the table.

Metric type ^Launch stage Display name
Kind, Type, Unit Monitored resources	Description Labels
smarthome_action/num_active_users ^BETA Daily active users
GAUGE, INT64, 1 assistant_action_project	Daily active users count. Sampled every 60 seconds. After sampling, data is not visible for up to 180 seconds. device_type: Device type, for example, "LIGHT" or "LOCK". locale: locale, for example, "en-US" or "fr-CA". trait: Assistant Smarthome trait, for example, "onOff" or "lockUnlock".
smarthome_action/request_count ^BETA Request count
DELTA, INT64, 1 assistant_action_project	Request count. Sampled every 60 seconds. After sampling, data is not visible for up to 180 seconds. device_type: Device type, for example, "LIGHT" or "LOCK". status: The canonical code that represents the result, for example, "SUCCESS" or "OFFLINE". trait: Assistant Smarthome trait, for example, "onOff" or "lockUnlock". is_local: (BOOL) Whether the request routed locally or through cloud endpoint. execution_type: Execution type, one of [execute, query].
smarthome_action/request_latencies ^BETA Request latencies
DELTA, DISTRIBUTION, ms assistant_action_project	Distribution of request latencies. Sampled every 60 seconds. After sampling, data is not visible for up to 180 seconds. device_type: Device type, for example, "LIGHT" or "LOCK". status: The canonical code that represents the result, for example, "SUCCESS" or "OFFLINE". trait: Assistant Smarthome trait, for example, "onOff" or "lockUnlock". is_local: (BOOL) Whether the request routed locally or through cloud endpoint. execution_type: Execution type, one of [execute, query].
smarthome_action/seven_day_active_users ^BETA 7 day active users
GAUGE, INT64, 1 assistant_action_project	7 day active users count. Sampled every 60 seconds. After sampling, data is not visible for up to 180 seconds. device_type: Device type, for example, "LIGHT" or "LOCK". locale: locale, for example, "en-US" or "fr-CA". trait: Assistant Smarthome trait, for example, "onOff" or "lockUnlock".
smarthome_action/twenty_eight_day_active_users ^BETA 28 day active users
GAUGE, INT64, 1 assistant_action_project	28 day active users count. Sampled every 60 seconds. After sampling, data is not visible for up to 180 seconds. device_type: Device type, for example, "LIGHT" or "LOCK". locale: locale, for example, "en-US" or "fr-CA". trait: Assistant Smarthome trait, for example, "onOff" or "lockUnlock".

aiplatform

Metrics from AI Platform (Unified). Launch stages of these metrics: BETA

The "metric type" strings in this table must be prefixed with aiplatform.googleapis.com/. That prefix has been omitted from the entries in the table.

Metric type ^Launch stage Display name
Kind, Type, Unit Monitored resources	Description Labels
prediction/online/accelerator/duty_cycle ^BETA Accelerator duty cycle
GAUGE, DOUBLE, 10^2.% aiplatform.googleapis.com/Endpoint	Average fraction of time over the past sample period during which the accelerator(s) were actively processing. Sampled every 60 seconds. After sampling, data is not visible for up to 360 seconds. deployed_model_id: The ID of the deployed model which serves the prediction request. replica_id: Unique ID corresponding to the deployed model replica.
prediction/online/accelerator/memory/bytes_used ^BETA Accelerator memory usage
GAUGE, INT64, By aiplatform.googleapis.com/Endpoint	Amount of accelerator memory allocated by the deployed model replica. Sampled every 60 seconds. After sampling, data is not visible for up to 360 seconds. deployed_model_id: The ID of the deployed model which serves the prediction request. replica_id: Unique ID corresponding to the deployed model replica.
prediction/online/cpu/utilization ^BETA CPU utilization
GAUGE, DOUBLE, 10^2.% aiplatform.googleapis.com/Endpoint	Fraction of CPU allocated by the deployed model replica and currently in use. May exceed 100% if the machine type has multiple CPUs. Sampled every 60 seconds. After sampling, data is not visible for up to 360 seconds. deployed_model_id: The ID of the deployed model which serves the prediction request. replica_id: Unique ID corresponding to the deployed model replica.
prediction/online/error_count ^BETA Error count
DELTA, INT64, 1 aiplatform.googleapis.com/Endpoint	Number of online prediction errors. deployed_model_id: The ID of the DeployedModel which serves the prediction request.
prediction/online/memory/bytes_used ^BETA Memory usage
GAUGE, INT64, By aiplatform.googleapis.com/Endpoint	Amount of memory allocated by the deployed model replica and currently in use. Sampled every 60 seconds. After sampling, data is not visible for up to 360 seconds. deployed_model_id: The ID of the deployed model which serves the prediction request. replica_id: Unique ID corresponding to the deployed model replica.
prediction/online/network/received_bytes_count ^BETA Network bytes received
DELTA, INT64, By aiplatform.googleapis.com/Endpoint	Number of bytes received over the network by the deployed model replica. Sampled every 60 seconds. After sampling, data is not visible for up to 360 seconds. deployed_model_id: The ID of the deployed model which serves the prediction request. replica_id: Unique ID corresponding to the deployed model replica.
prediction/online/network/sent_bytes_count ^BETA Network bytes sent
DELTA, INT64, By aiplatform.googleapis.com/Endpoint	Number of bytes sent over the network by the deployed model replica. Sampled every 60 seconds. After sampling, data is not visible for up to 360 seconds. deployed_model_id: The ID of the deployed model which serves the prediction request. replica_id: Unique ID corresponding to the deployed model replica.
prediction/online/prediction_count ^BETA Prediction count
DELTA, INT64, 1 aiplatform.googleapis.com/Endpoint	Number of online predictions. deployed_model_id: The ID of the DeployedModel which serves the prediction request.
prediction/online/prediction_latencies ^BETA Prediction Latencies
DELTA, DISTRIBUTION, ms aiplatform.googleapis.com/Endpoint	Online prediction latency of the deployed model. deployed_model_id: The ID of the DeployedModel which serves the prediction request. latency_type: One of the following "overhead", "model" or "total". The "total" latency type will always be populated. And the "overhead" and "model" latency type only get populated for custom-trained models and AutoML tabular models.
prediction/online/replicas ^BETA Replica count
GAUGE, INT64, 1 aiplatform.googleapis.com/Endpoint	Number of active replicas used by the deployed model. deployed_model_id: The ID of the DeployedModel which serves the prediction request.
prediction/online/response_count ^BETA Response count
DELTA, INT64, 1 aiplatform.googleapis.com/Endpoint	Number of different online prediction response codes. deployed_model_id: The ID of the DeployedModel which serves the prediction request. response_code: HTTP response code (1xx, 2xx, 3xx, 4xx, 5xx).
prediction/online/target_replicas ^BETA Replica target
GAUGE, INT64, 1 aiplatform.googleapis.com/Endpoint	Target number of active replicas needed for the deployed model. deployed_model_id: The ID of the DeployedModel which serves the prediction request.

apigateway

Metrics from API Gateway. Launch stages of these metrics: BETA

The "metric type" strings in this table must be prefixed with apigateway.googleapis.com/. That prefix has been omitted from the entries in the table.

Metric type ^Launch stage Display name
Kind, Type, Unit Monitored resources	Description Labels
proxy/request_count ^BETA Request Count
DELTA, INT64, 1 apigateway.googleapis.com/Gateway	Number of requests served by API gateway. response_code_class: HTTP response code class (1xx, 2xx, 3xx, 4xx, 5xx). api_config: The name of API Config resource hosted by the Gateway.

apigee

Metrics from Apigee. Launch stages of these metrics: ALPHA BETA

The "metric type" strings in this table must be prefixed with apigee.googleapis.com/. That prefix has been omitted from the entries in the table.

Metric type ^Launch stage Display name
Kind, Type, Unit Monitored resources	Description Labels
cassandra/clientrequest_latency ^BETA Apigee Cassandra client request latency
GAUGE, DOUBLE, 1 k8s_container	Average read request latency in the 75th percentile range in microseconds for Apigee Cassandra. scope: Scope of the request. unit: Unit of the metric.
cassandra/compaction_pendingtasks ^BETA Apigee Cassandra pending compaction tasks
GAUGE, INT64, 1 k8s_container	Number of pending compaction tasks for Apigee Cassandra. unit: unit of the metric.
cassandra/jvm_memory_bytes_committed ^BETA Apigee Cassandra bytes committed per area
GAUGE, DOUBLE, By k8s_container	Number of bytes committed for Apigee Cassandra JVM heap. area: Area of the memory such as heap or non-heap.
cassandra/jvm_memory_bytes_init ^BETA Apigee Cassandra initial memory bytes
GAUGE, DOUBLE, By k8s_container	Initial memory usage in bytes for Apigee Cassandra JVM heap. area: Area of the memory such as heap or non-heap.
cassandra/jvm_memory_bytes_max ^BETA Apigee Cassandra max bytes of memory
GAUGE, DOUBLE, By k8s_container	Maximum memory usage in bytes for Apigee Cassandra JVM heap. area: Area of the memory such as heap or non-heap.
cassandra/jvm_memory_bytes_used ^BETA Apigee Cassandra used JVM memory bytes
GAUGE, DOUBLE, By k8s_container	Memory usage in bytes for Apigee Cassandra JVM heap. area: Area of the memory such as heap or non-heap.
cassandra/jvm_memory_pool_bytes_committed ^BETA Apigee Cassandra bytes committed per memory pool
GAUGE, DOUBLE, By k8s_container	Memory usage in bytes committed per Apigee Cassandra JVM pool. pool: Name of the cassandra memory pool.
cassandra/jvm_memory_pool_bytes_init ^BETA Apigee Cassandra initial bytes of JVM memory pool
GAUGE, DOUBLE, By k8s_container	Initial memory usage in bytes for Apigee Cassandra JVM pool. pool: Name of the cassandra memory pool.
cassandra/jvm_memory_pool_bytes_max ^BETA Apigee Cassandra JVM memory pool bytes max
GAUGE, DOUBLE, By k8s_container	Maximum memory usage in bytes for Apigee Cassandra JVM pool. pool: Name of the cassandra memory pool.
cassandra/jvm_memory_pool_bytes_used ^BETA Apigee Cassandra bytes per memory pool
GAUGE, DOUBLE, By k8s_container	Memory usage in bytes per Apigee Cassandra JVM pool. pool: Name of the cassandra memory pool.
cassandra/process_cpu_seconds_total ^BETA Apigee Cassandra user and system CPU in seconds
CUMULATIVE, DOUBLE, s k8s_container	User and system CPU time spent in seconds for Apigee Cassandra.
cassandra/process_max_fds ^BETA Apigee Cassandra process max file descriptors
GAUGE, INT64, 1 k8s_container	Maximum number of open file descriptors for Apigee Cassandra.
cassandra/process_open_fds ^BETA Apigee Cassandra process open file descriptors
GAUGE, INT64, 1 k8s_container	Number of open file descriptors for Apigee Cassandra.
connectagent/completed_rpcs_count ^ALPHA Apigee Connect Agent completed RPCs count
CUMULATIVE, INT64, 1 k8s_container	Count of RPCs completed by Apigee Connect Agent. method: GRPC client method. status: GRPC client Status.
connectagent/dropped_requests ^ALPHA Apigee Connect Agent dropped requests
CUMULATIVE, DISTRIBUTION, 1 k8s_container	Distribution of dropped requests by Apigee Connect Agent per RPC. method: GRPC client method.
connectagent/dropped_responses ^ALPHA Apigee Connect Agent dropped responses
CUMULATIVE, DISTRIBUTION, 1 k8s_container	Distribution of dropped responses by Apigee Connect Agent per RPC. method: GRPC client method.
connectagent/received_bytes ^ALPHA Apigee Connect Agent received bytes
CUMULATIVE, DISTRIBUTION, By k8s_container	Distribution of bytes received by Apigee Connect Agent per RPC. method: GRPC client method.
connectagent/received_messages ^ALPHA Apigee Connect Agent received messages
CUMULATIVE, DISTRIBUTION, 1 k8s_container	Distribution of messages received by Apigee Connect Agent per RPC. method: GRPC client method.
connectagent/sent_bytes ^ALPHA Apigee Connect Agent sent bytes
CUMULATIVE, DISTRIBUTION, By k8s_container	Distribution of bytes sent by Apigee Connect Agent per RPC. method: GRPC client method.
connectagent/sent_messages ^ALPHA Apigee Connect Agent sent messages
CUMULATIVE, DISTRIBUTION, 1 k8s_container	Distribution of messages sent by Apigee Connect Agent per RPC. method: GRPC client method.
connectagent/stream_liveness_time ^ALPHA Apigee Connect Agent stream liveness time
CUMULATIVE, DISTRIBUTION, ms k8s_container	Distribution of stream liveness time by Apigee Connect Agent per RPC. method: GRPC client method.
environment/anomaly_count ^BETA Apigee anomaly event count
CUMULATIVE, INT64, 1 apigee.googleapis.com/Environment	Total number of anomalies observed in the Apigee environment. severity: Anomaly severity (slight, moderate, severe). direction: Anomaly direction (jump/drop). anomaly_type: Type of Apigee anomaly (example: traffic_2xx).
policy/latencies ^BETA Apigee policy response latencies
CUMULATIVE, DISTRIBUTION, ms apigee.googleapis.com/Proxy	Total amount of time in milliseconds for an Apigee policy to execute. policy_name: Name of the policy.
policyv2/latencies_percentile ^BETA Percentile of Apigee policy response latencies
GAUGE, DOUBLE, ms apigee.googleapis.com/ProxyV2	Percentile of all API policy responseto a request. policy_name: Name of the policy. percentile: Apigee policy response latencies Percentile. time_window: Apigee policy response latencies percentile aggregation time window.
proxy/latencies ^BETA Apigee proxy response latencies
CUMULATIVE, DISTRIBUTION, ms apigee.googleapis.com/Proxy	Total amount of time in milliseconds for the API proxy to respond to a request. This time includes the Apigee API proxy overhead and the target server time. method: HTTP method.
proxy/request_count ^BETA Apigee proxy request cumulative count
CUMULATIVE, INT64, 1 apigee.googleapis.com/Proxy	Total number of requests received by the Apigee API proxy. method: HTTP method.
proxy/response_count ^BETA Apigee proxy response cumulative count
CUMULATIVE, INT64, 1 apigee.googleapis.com/Proxy	Total number of responses sent by the Apigee API proxy. method: HTTP method. response_code: HTTP response code.
proxyv2/latencies_percentile ^BETA Percentile of Apigee proxy response latencies
GAUGE, DOUBLE, ms apigee.googleapis.com/ProxyV2	Percentile of all API proxy responseto a request. method: HTTP method. percentile: Apigee proxy response latencies Percentile. time_window: Apigee Proxy response latencies percentile aggregation time window.
proxyv2/request_count ^ALPHA Apigee proxy request cumulative count
CUMULATIVE, INT64, 1 apigee.googleapis.com/ProxyV2	Total number of requests received by the Apigee API proxy. method: HTTP method.
proxyv2/response_count ^ALPHA Apigee proxy response cumulative count
CUMULATIVE, INT64, 1 apigee.googleapis.com/ProxyV2	Total number of responses sent by the Apigee API proxy. method: HTTP method. response_code: HTTP response code. fault_code: Apigee fault code that provides useful information to help you diagnose issues. fault_source: Information about the suspected cause of the fault, such as API proxy, target, or the Apigee infrastructure.
server/fault_count ^BETA Apigee server fault count
CUMULATIVE, INT64, 1 k8s_container	Total number of faults returned by the Apigee server application. For example, the apigee-runtime, apigee--synchronizer, or apigee-udca server application. source: Source of the fault.
server/latencies ^BETA Apigee server latencies
CUMULATIVE, DISTRIBUTION, ms k8s_container	Total amount of time in milliseconds for the Apigee application server to process the requests and responses. For example, the apigee-runtime, apigee-synchronizer, or apigee-udca server application. method: HTTP method. response_code: HTTP response code. type: Type of service.
server/nio ^BETA Apigee server nio
GAUGE, INT64, 1 k8s_container	Number of open sockets. state: State of the NIO threads.
server/num_threads ^BETA Apigee server thread count
GAUGE, INT64, 1 k8s_container	Number of active non-daemon threads in the Apigee server.
server/request_count ^BETA Apigee server request count
CUMULATIVE, INT64, 1 k8s_container	Total number of requests received by the Apigee server application. For example, the apigee-runtime, apigee-synchronizer, or apigee-udca server application. method: HTTP method. type: Type of service.
server/response_count ^BETA Apigee server response count
CUMULATIVE, INT64, 1 k8s_container	Total number of responses sent by the Apigee server application. For example, the apigee-runtime, apigee-synchronizer, or apigee-udca server application. method: HTTP method. response_code: HTTP response code. type: Type of service.
target/latencies ^ALPHA Apigee target response latencies
CUMULATIVE, DISTRIBUTION, ms apigee.googleapis.com/Proxy	Total amount of time in milliseconds for the Apigee API proxy target to respond to a request. This value does not include the Apigee API proxy overhead. method: HTTP method. target_type: Type of target such as service callout. target_endpoint: Target endpoint URL. response_code: HTTP response code.
target/request_count ^ALPHA Apigee target request cumulative count
CUMULATIVE, INT64, 1 apigee.googleapis.com/Proxy	Total number of requests received by the Apigee API proxy target. method: HTTP method. target_type: Type of target such as service callout. target_endpoint: Target endpoint URL.
target/response_count ^ALPHA Apigee target response cumulative count
CUMULATIVE, INT64, 1 apigee.googleapis.com/Proxy	Total number of responses sent by the Apigee API proxy target. method: HTTP method. target_type: Type of target such as service callout. target_endpoint: Target endpoint URL. response_code: HTTP response code.
targetv2/request_count ^ALPHA Apigee target request cumulative count
CUMULATIVE, INT64, 1 apigee.googleapis.com/ProxyV2	Total number of requests received by the Apigee API proxy target. method: HTTP method.
targetv2/response_count ^ALPHA Apigee target response cumulative count
CUMULATIVE, INT64, 1 apigee.googleapis.com/ProxyV2	Total number of responses sent by the Apigee API proxy target. method: HTTP method. response_code: HTTP response code.
udca/disk/used_bytes ^BETA Apigee UDCA disk used bytes
GAUGE, DOUBLE, By k8s_container	Amount of bytes used by the data files on the data collection pod's disk. dataset: Type of data. state: State of the data file.
udca/server/local_file_count ^BETA Apigee UDCA server local file count
GAUGE, INT64, 1 k8s_container	Number of files on disk in the data collection pod. Ideally, the value will be close to 0. A consistently high value indicates that files are not being uploaded or that the Apigee Universal Data Collection Agent (UDCA) server application is not able to upload them fast enough. This value is computed every 60 seconds and does not reflect the state of the Apigee UDCA server application in real time. dataset: Type of data. state: State of the data file.
udca/server/local_file_latest_ts ^BETA Apigee UDCA server timestamp difference between current time and latest file
GAUGE, INT64, s k8s_container	Apigee Universal Data Collection Agent (UDCA) server timestamp difference between current time and latest file. dataset: Type of data. state: State of the data file.
udca/server/local_file_oldest_ts ^BETA Apigee UDCA server timestamp difference between current time and oldest file
GAUGE, INT64, s k8s_container	Time that the oldest file in the dataset was created in seconds since epoch. If the Apigee Universal Data Collection Agent (UDCA) server application is up to date and there are no files waiting to be uploaded when this metric is computed, then this value will be 0. If this value increases, it indicates that old files are still on disk. The metric is computed every 60 seconds and does not reflect the state of the Apigee UDCA server application in real time. dataset: Type of data. state: State of the data file.
udca/server/pruned_file_count ^BETA Apigee UDCA pruned file count
CUMULATIVE, INT64, 1 k8s_container	Total number of files that have been deleted because their Time To Life (TTL) was beyond a set threshold. The dataset can include API, trace, and others, and state can be UPLOADED, FAILED, or DISCARDED. dataset: Type of data. state: State of the data file.
udca/server/retry_cache_size ^BETA Apigee UDCA outstanding number of entries in retry cache
GAUGE, INT64, 1 k8s_container	Number of files, by dataset, that the Apigee Universal Data Collection Agent (UDCA) server application is retrying to upload. After 3 retries per file, the UDCA moves the file to the /failed subdirectory and removes it from this cache. An increase in this value over time implies that the cache is not being cleared. dataset: Type of data.
udca/server/total_latencies ^BETA Apigee UDCA server total latencies
CUMULATIVE, DISTRIBUTION, s k8s_container	Total amount of time in seconds for a data file to be created and uploaded successfully by the Apigee Universal Data Collection Agent (UDCA) server application. Buckets will be 100 ms, 250 ms, 500 ms, 1 s, 2 s, 4 s, 8 s, 16 s, 32 s, and 64 s. The metrics will display a histogram for total latency from file creation time to successful upload time. dataset: Type of data.
udca/server/upload_latencies ^BETA Apigee UDCA server upload latencies
CUMULATIVE, DISTRIBUTION, s k8s_container	Total amount of time in seconds that the Apigee Universal Data Collection Agent (UDCA) server application required to upload a data file. Buckets will be 100 ms, 250 ms, 500 ms, 1 s, 2 s, 4 s, 8 s500 ms, 1 s, 2 s, 4 s, 8 s, 16 s, 32 s, and 64 s. The metrics will display a histogram for total latency from file creation time to successful upload time. dataset: Type of data.
udca/upstream/http_error_count ^BETA Apigee UDCA server HTTP error count
CUMULATIVE, INT64, 1 k8s_container	Total number of HTTP errors encountered by the Apigee Universal Data Collection Agent (UDCA) server application. This metric is useful to help determin in which part of the UDCA external dependencies are failing and for what reason. These errors can arise for various services (such as getDataLocation, Cloud storage, token generator) and for various datasets (such as API and trace) with different response codes. dataset: Type of data. service: Upstream service name. response_code: HTTP response code.
udca/upstream/http_latencies ^BETA Apigee UDCA server HTTP latencies
CUMULATIVE, DISTRIBUTION, s k8s_container	Total amount of time in seconds required for upstream services. Buckets will be 100 ms, 250 ms, 500 ms, 1 s, 2 s, 4 s, 8 s, 16 s, 32 s, and 64 s. The metrics will display a histogram for latency from upstream services. dataset: Type of data. service: Upstream service name.
udca/upstream/uploaded_file_count ^BETA Apigee UDCA uploaded file count
CUMULATIVE, INT64, 1 k8s_container	Total number of files being uploaded to the Apigee Universal Data Collection Agent (UDCA) server application. Note that the event dataset value should keep growing, the API dataset value should keep growing if the organization and environment have constant traffic, and the trace dataset value should increase when you use the Apigee trace tools to debug or inspect your requests. dataset: Type of data.
udca/upstream/uploaded_file_sizes ^BETA Apigee UDCA uploaded file size distribution
CUMULATIVE, DISTRIBUTION, By k8s_container	Total file size in bytes uploaded to the Apigee Universal Data Collection Agent (UDCA) server application. Buckets will be 1 KB, 10 KB, 100 KB, 1 MB, 10 MB, 100 MB, and 1 GB. The metrics will display a histogram for file size by dataset, organization, and environment. dataset: Type of data.
upstream/latencies ^BETA Apigee upstream latencies
CUMULATIVE, DISTRIBUTION, ms k8s_container	Total amount of time in milliseconds incurred by the upstream Apigee server application. For apigee-synchronizer this metric indicates the latency from the control plane. method: HTTP method. response_code: HTTP response code. type: Type of service.
upstream/request_count ^BETA Apigee upstream request count
CUMULATIVE, INT64, 1 k8s_container	Total number of requests sent by the Apigee server application to its upstream application. For example, for the apigee-synchronizer the control plane is upstream. method: HTTP method. type: Type of service.
upstream/response_count ^BETA Apigee upstream response count
CUMULATIVE, INT64, 1 k8s_container	Total number of responses received by the Apigee server application from its upstream application. For example, for the apigee-synchronizer the control plane is upstream. method: HTTP method. response_code: HTTP response code. type: Type of service.

appengine

Metrics from App Engine. Launch stages of these metrics: ALPHA BETA GA

The "metric type" strings in this table must be prefixed with appengine.googleapis.com/. That prefix has been omitted from the entries in the table.

Metric type ^Launch stage Display name
Kind, Type, Unit Monitored resources	Description Labels
flex/autoscaler/capacity ^BETA Autoscaling Metrics Utilization Capacity
GAUGE, DOUBLE, 1 gae_app	Utilization capacity multiplied by number of serving VMs. Sampled every 60 seconds. After sampling, data is not visible for up to 240 seconds.
flex/autoscaler/connections/current ^BETA Writing and reading connections
GAUGE, DOUBLE, 1 gce_instance	Number of current reading and writing connections per App Engine flexible environment instance. Intended to be used for autoscaling. Sampled every 60 seconds. After sampling, data is not visible for up to 120 seconds. service_id: The service/module name. state: State of the connection (writing or reading). version_id: The version name. instance_name: The VM instance name.
flex/autoscaler/current_utilization ^BETA Autoscaling Metrics Current Utilization
GAUGE, DOUBLE, 1 gae_app	The sum of the utilization of a specified metric for all serving VMs. Sampled every 60 seconds. After sampling, data is not visible for up to 240 seconds. metric_type: The target autoscaling metric type. For example, cpu.
flex/autoscaler/server/request_count ^BETA Request counts
DELTA, INT64, 1 gce_instance	Request counts for an App Engine flexible environment instance.Intended to be used for autoscaling. Sampled every 60 seconds. After sampling, data is not visible for up to 240 seconds. service_id: The service/module name. version_id: The version name. instance_name: The VM instance name.
flex/connections/current ^GA Connections
GAUGE, DOUBLE, 1 gae_app	Number of current active connections per App Engine flexible environment version. Sampled every 60 seconds. After sampling, data is not visible for up to 240 seconds.
flex/cpu/reserved_cores ^GA Reserved cores
GAUGE, DOUBLE, 1 gae_app	Total number of CPU cores allocated to an App Engine flexible environment version. Sampled every 60 seconds. After sampling, data is not visible for up to 240 seconds.
flex/cpu/utilization ^GA CPU utilization
GAUGE, DOUBLE, 10^2.% gae_app	Fractional utilization of allocated CPU across an AppEngine flexible environment version. Values are typically numbers between 0.0 and 1.0 (but some machine types allow bursting above 1.0). Charts display the values as a percentage between 0% and 100% (or more). Sampled every 60 seconds. After sampling, data is not visible for up to 240 seconds.
flex/disk/read_bytes_count ^GA Disk bytes read
DELTA, INT64, By gae_app	Delta count of bytes read from disk across an App Engine flexible environment version. Sampled every 60 seconds. After sampling, data is not visible for up to 240 seconds.
flex/disk/write_bytes_count ^GA Disk bytes written
DELTA, INT64, By gae_app	Delta count of bytes written from disk across an App Engine flexible environment version. Sampled every 60 seconds. After sampling, data is not visible for up to 240 seconds.
flex/instance/connections/current ^GA Connections
GAUGE, DOUBLE, 1 gae_instance	Number of current active connections per App Engine flexible environment instance. Sampled every 60 seconds. After sampling, data is not visible for up to 240 seconds.
flex/instance/cpu/usage_time ^ALPHA CPU Usage
DELTA, DOUBLE, s{CPU} gae_instance	Delta vCPU usage for all vCPUs, in vCPU-seconds. This value is reported by the hypervisor for the VM and can differ from `agent.googleapis.com/cpu/usage_time`, which is reported from inside the VM. Sampled every 60 seconds. After sampling, data is not visible for up to 240 seconds.
flex/instance/cpu/utilization ^GA CPU Utilization
GAUGE, DOUBLE, 10^2.% gae_instance	Fractional CPU utilization for all cores on a single AppEngine flexible instance. Values are typically numbers between 0.0 and 1.0 for a single core (but might exceed 1.0 in total). Charts display the values as a percentage between 0% and 100% (or more). Sampled every 60 seconds. After sampling, data is not visible for up to 240 seconds.
flex/instance/disk/io/read_bytes_count ^ALPHA Disk read bytes
DELTA, INT64, By gae_instance	Delta count of bytes read from disk. Sampled every 60 seconds. After sampling, data is not visible for up to 240 seconds.
flex/instance/disk/io/write_bytes_count ^ALPHA Disk write bytes
DELTA, INT64, By gae_instance	Delta count of bytes written to disk. Sampled every 60 seconds. After sampling, data is not visible for up to 240 seconds.
flex/instance/guest/disk/bytes_used ^ALPHA Disk usage
GAUGE, INT64, By gae_instance	Number of bytes used on disk for file systems. Sampled every 60 seconds. After sampling, data is not visible for up to 240 seconds. device_name: The name of the disk device. state: Type of usage (free, used, reserved).
flex/instance/guest/memory/bytes_used ^ALPHA Memory usage
GAUGE, INT64, By gae_instance	Memory usage by each memory state, in Bytes. Summing values of all states yields the total memory on the machine. Sampled every 60 seconds. After sampling, data is not visible for up to 240 seconds. state: One of the following: buffered, cached, free, slab, used.
flex/instance/healthcheck/liveness/request_count ^ALPHA Liveness check
DELTA, INT64, 1 gae_instance	Number of liveness health-check requests for a Flex instance with status and error code. Sampled every 60 seconds. After sampling, data is not visible for up to 240 seconds. status: Status of healthcheck request. error_code: Error code when status is 'fail'.
flex/instance/healthcheck/readiness/request_count ^ALPHA Readiness check
DELTA, INT64, 1 gae_instance	Number of readiness health-check requests for a Flex instance with status and error code. Sampled every 60 seconds. After sampling, data is not visible for up to 240 seconds. status: Status of healthcheck request. error_code: Error code when status is 'fail'.
flex/instance/jvm/gc/count ^ALPHA GC count
DELTA, INT64, 1 gae_instance	The total number of garbage collections that have occurred. Sampled every 60 seconds. Sampled every 60 seconds. After sampling, data is not visible for up to 240 seconds. gc_name: The name of the garbage collector involved.
flex/instance/jvm/gc/time ^ALPHA GC Time
DELTA, INT64, ms gae_instance	The accumulated garbage collection elapsed time in milliseconds. Sampled every 60 seconds. Sampled every 60 seconds. After sampling, data is not visible for up to 240 seconds. gc_name: The name of the garbage collector involved.
flex/instance/log_entry_count ^ALPHA Logging Agent Log Entry Count
DELTA, INT64, 1 gae_instance	Count of log entries written by the Logging agent. Sampled every 60 seconds. Sampled every 60 seconds. After sampling, data is not visible for up to 240 seconds. response_code: (INT64) Numeric response code from both the gRPC and HTTP error spaces.
flex/instance/log_entry_retry_count ^ALPHA Logging Agent Retried Log Entry Writes Count
DELTA, INT64, 1 gae_instance	Count of log-entry writes that were retried by the Logging agent. Sampled every 60 seconds. Sampled every 60 seconds. After sampling, data is not visible for up to 240 seconds. response_code: (INT64) Numeric response code from both the gRPC and HTTP error spaces.
flex/instance/network/received_bytes_count ^GA Network bytes received
DELTA, INT64, By gae_instance	Delta count of incoming network bytes in an App Engine flexible instance. Sampled every 60 seconds. After sampling, data is not visible for up to 240 seconds.
flex/instance/network/sent_bytes_count ^GA Network bytes sent
DELTA, INT64, By gae_instance	Delta count of outgoing network bytes in an App Engine flexible instance. Sampled every 60 seconds. After sampling, data is not visible for up to 240 seconds.
flex/instance/nginx/connections/accepted_count ^ALPHA Accepted connections
DELTA, INT64, 1 gae_instance	Total connections accepted. Sampled every 60 seconds. After sampling, data is not visible for up to 240 seconds.
flex/instance/nginx/connections/current ^ALPHA Active connections
GAUGE, DOUBLE, 1 gae_instance	The number of connections currently attached to Nginx. Sampled every 60 seconds. After sampling, data is not visible for up to 240 seconds. state: The Nginx connection state.
flex/instance/nginx/connections/handled_count ^ALPHA Handled connections
DELTA, INT64, 1 gae_instance	Total connections handled. Sampled every 60 seconds. After sampling, data is not visible for up to 240 seconds.
flex/instance/uptime ^ALPHA Uptime
GAUGE, DOUBLE, s gae_instance	How long the VM has been running, in seconds. Sampled every 60 seconds. After sampling, data is not visible for up to 240 seconds.
flex/instance/ws/avg_duration ^GA Websocket average duration
GAUGE, DOUBLE, s gae_instance	Average of websocket requests for a given App Engine Flexible instance. Sampled every 60 seconds. After sampling, data is not visible for up to 240 seconds.
flex/network/received_bytes_count ^GA Network bytes received.
DELTA, INT64, By gae_app	Delta count of incoming network bytes across all VMs in an App Engine flexible environment version. Sampled every 60 seconds. After sampling, data is not visible for up to 240 seconds.
flex/network/sent_bytes_count ^GA Network bytes sent.
DELTA, INT64, By gae_app	Delta count of outgoing network bytes across all VMs in an App Engine flexible environment version. Sampled every 60 seconds. After sampling, data is not visible for up to 240 seconds.
http/server/dos_intercept_count ^GA Interception count
DELTA, INT64, 1 gae_app	Delta count of interceptions performed to prevent DoS attacks. Sampled every 60 seconds. After sampling, data is not visible for up to 240 seconds.
http/server/quota_denial_count ^GA Quota denial count
DELTA, INT64, 1 gae_app	Delta count of requests that failed due to the app being over quota. Sampled every 60 seconds. After sampling, data is not visible for up to 240 seconds.
http/server/response_count ^GA Response count
DELTA, INT64, 1 gae_app	Delta HTTP response count. Sampled every 60 seconds. After sampling, data is not visible for up to 240 seconds. loading: (BOOL) Indicates whether the request was loading. response_code: (INT64) HTTP status code for the request.
http/server/response_latencies ^GA Response latency
DELTA, DISTRIBUTION, ms gae_app	HTTP response latency. Sampled every 60 seconds. After sampling, data is not visible for up to 240 seconds. loading: (BOOL) Indicates whether the request was loading. response_code: (INT64) HTTP status code for the request.
http/server/response_style_count ^GA Response count by style
DELTA, INT64, 1 gae_app	Delta counts on the HTTP serve style. Sampled every 60 seconds. After sampling, data is not visible for up to 240 seconds. dynamic: (BOOL) Indicates if the request was a static or dynamic request. cached: (BOOL) Indicates if the response was cached.
memcache/centi_mcu_count ^GA Memcache utilization
DELTA, INT64, 1 gae_app	Memcache utilization in one hundredth of Memcache Compute Unit grouped by command. Sampled every 60 seconds. After sampling, data is not visible for up to 240 seconds. command: Name of the memcache command.
memcache/hit_ratio ^GA Hit ratio
GAUGE, DOUBLE, 1 gae_app	Percentage of successful hits compared to all memcache operations. Sampled every 60 seconds. After sampling, data is not visible for up to 240 seconds.
memcache/operation_count ^GA Memcache operations
DELTA, INT64, 1 gae_app	Count of memcache key operations, grouped by command and status. Sampled every 60 seconds. After sampling, data is not visible for up to 180 seconds. command: Name of the memcache command. status: Whether the operation succeeded or not, e.g. HIT or MISS. This is currently unpopulated (always an empty string).
memcache/received_bytes_count ^GA Memcache received bytes
DELTA, INT64, By gae_app	Number of bytes received by app from the memcache API, grouped by status and memcache command. Sampled every 60 seconds. After sampling, data is not visible for up to 240 seconds. status: RPC status of the request. command: Name of the memcache command.
memcache/sent_bytes_count ^GA Memcache sent bytes
DELTA, INT64, By gae_app	Number of bytes sent by app through the memcache API, grouped by memcache command. Sampled every 60 seconds. After sampling, data is not visible for up to 240 seconds. command: Name of the memcache command.
memcache/used_cache_size ^GA Used Cache Size
GAUGE, INT64, By gae_app	Size of the cache, calculated as the total size of all items stored. Sampled every 60 seconds. After sampling, data is not visible for up to 240 seconds.
system/billed_instance_estimate_count ^GA Estimated instance count
GAUGE, INT64, 1 gae_app	Estimate of the number of instances that the customer is charged for. Sampled every 60 seconds. After sampling, data is not visible for up to 240 seconds.
system/cpu/usage ^GA CPU megacycles
GAUGE, INT64, 1 gae_app	CPU usage in megacycles over all instances. Sampled every 60 seconds. After sampling, data is not visible for up to 240 seconds. source: The source of the CPU usage (api or runtime).
system/instance_count ^GA Instance count
GAUGE, INT64, 1 gae_app	Number of instances that exist. Sampled every 60 seconds. After sampling, data is not visible for up to 240 seconds. state: State of instance (active, idle, etc).
system/memory/usage ^GA Memory usage
GAUGE, INT64, By gae_app	Total memory used by running instances. Sampled every 60 seconds. After sampling, data is not visible for up to 240 seconds.
system/network/received_bytes_count ^GA Received bytes
DELTA, INT64, By gae_app	Delta count of incoming network bandwidth. Sampled every 60 seconds. After sampling, data is not visible for up to 240 seconds. cached: (BOOL) Indicates if response was cached.
system/network/sent_bytes_count ^GA Sent bytes
DELTA, INT64, By gae_app	Delta count of outgoing network bandwidth. Sampled every 60 seconds. After sampling, data is not visible for up to 240 seconds. cached: (BOOL) Indicates if response was cached.

autoscaler

Metrics from Compute Engine Autoscaler. Launch stages of these metrics: BETA

The "metric type" strings in this table must be prefixed with autoscaler.googleapis.com/. That prefix has been omitted from the entries in the table.

Metric type ^Launch stage Display name
Kind, Type, Unit Monitored resources	Description Labels
capacity ^BETA Serving capacity
GAUGE, DOUBLE, 1 autoscaler	Utilization target multiplied by number of serving VMs. Sampled every 60 seconds. After sampling, data is not visible for up to 180 seconds. metric_type: Name of the resource the autoscaling is based on - "cpu", "load balancing" or full metric name in domain-and-path format if scaling is based on Cloud Monitoring metric.
current_utilization ^BETA Current Autoscaler utilization
GAUGE, DOUBLE, 1 autoscaler	The sum of the utilization of a specified metric for all serving VMs. Sampled every 60 seconds. After sampling, data is not visible for up to 180 seconds. metric_type: Name of the resource the autoscaling is based on - "cpu", "load balancing" or full metric name in domain-and-path format if scaling is based on Cloud Monitoring metric.

bigquery

Metrics from BigQuery.

Note: When a job runs in less than 60 seconds, metrics sampled every 60 seconds might not report any data.

Launch stages of these metrics: BETA EARLY_ACCESS GA

The "metric type" strings in this table must be prefixed with bigquery.googleapis.com/. That prefix has been omitted from the entries in the table.

Metric type ^Launch stage Display name
Kind, Type, Unit Monitored resources	Description Labels
job/num_in_flight ^GA Job count
GAUGE, INT64, 1 bigquery_project	In flight jobs. Sampled every 60 seconds. After sampling, data is not visible for up to 420 seconds. priority: Job priority (batch or interactive). job_type: Job type. state: Job state.
query/count ^GA Query count
GAUGE, INT64, 1 bigquery_project, global	In flight queries. Sampled every 60 seconds. After sampling, data is not visible for up to 420 seconds. priority: Query priority (batch or interactive).
query/execution_times ^GA Query execution times
GAUGE, DISTRIBUTION, s bigquery_project, global	Distribution of queries execution times. Sampled every 60 seconds. After sampling, data is not visible for up to 420 seconds. priority: Query priority (batch or interactive).
query/scanned_bytes ^GA Scanned bytes
DELTA, INT64, By global	Scanned bytes. Sampled every 60 seconds. After sampling, data is not visible for up to 21720 seconds. priority: Query priority (batch or interactive).
query/scanned_bytes_billed ^GA Scanned bytes billed
DELTA, INT64, By global	Scanned bytes billed. Sampled every 60 seconds. After sampling, data is not visible for up to 21720 seconds. priority: Query priority (batch or interactive).
query/statement_scanned_bytes ^BETA Statement scanned bytes
DELTA, INT64, By bigquery_project	Scanned bytes broken down by statement type. Sampled every 60 seconds. After sampling, data is not visible for up to 21720 seconds. priority: Query priority (batch or interactive). statement_type: Query statement type (SELECT, CREATE_MODEL ...).
query/statement_scanned_bytes_billed ^BETA Statement scanned bytes billed
DELTA, INT64, By bigquery_project	Scanned bytes billed broken down by statement type. Sampled every 60 seconds. After sampling, data is not visible for up to 21720 seconds. priority: Query priority (batch or interactive). statement_type: Query statement type (SELECT, CREATE_MODEL ...).
slots/allocated ^GA Slots used by project, reservation, and job type
GAUGE, INT64, 1 bigquery_project, global	Number of BigQuery slots currently allocated for project. Slot allocation can be broken down based on reservation and job type. Sampled every 60 seconds. After sampling, data is not visible for up to 420 seconds. reservation: Reservation. This field is unset if the query is on-demand. job_type: Job type.
slots/allocated_for_project ^GA Slots used by project
GAUGE, INT64, 1 global	Number of BigQuery slots currently allocated for query jobs in the project. Sampled every 60 seconds. After sampling, data is not visible for up to 420 seconds.
slots/allocated_for_project_and_job_type ^GA Slots used by project and job type
GAUGE, INT64, 1 global	Number of BigQuery slots currently allocated for the project and job type. Sampled every 60 seconds. After sampling, data is not visible for up to 420 seconds. job_type: Job type.
slots/allocated_for_reservation ^GA Slots used by project in reservation
GAUGE, INT64, 1 global	Number of BigQuery slots currently allocated for project in the reservation. Sampled every 60 seconds. After sampling, data is not visible for up to 420 seconds.
slots/assigned ^EARLY_ACCESS Slots assigned
GAUGE, INT64, 1 bigquery_project	The number of slots assigned to the given project/organization. Sampled every 60 seconds. After sampling, data is not visible for up to 120 seconds. job_type: Job type. reservation: Reservation.
slots/capacity_committed ^EARLY_ACCESS Slots capacity committed
GAUGE, INT64, 1 bigquery_project	The total slot capacity commitments purchased through this administrator project/organization. Sampled every 60 seconds. After sampling, data is not visible for up to 120 seconds. plan: Capacity commitment plan.
slots/max_assigned ^EARLY_ACCESS Slots max assigned
GAUGE, INT64, 1 bigquery_project	The maximum number of slots assigned to the given project/organization. Sampled every 60 seconds. After sampling, data is not visible for up to 120 seconds. job_type: Job type. reservation: Reservation.
slots/total_allocated_for_reservation ^GA Slots used across projects in reservation
GAUGE, INT64, 1 bigquery_project, global	Number of BigQuery slots currently allocated across projects in the reservation. This is only reported while the project is using slots in the reservation. Sampled every 60 seconds. After sampling, data is not visible for up to 420 seconds.
slots/total_available ^DEPRECATED Total slots
GAUGE, INT64, 1 bigquery_project, global	(Deprecated) Total number of BigQuery slots available for the project. If you are using the BigQuery Reservations API, consider using bigquery.googleapis.com/slots/assigned and bigquery.googleapis.com/slots/max_assigned instead. Sampled every 60 seconds. After sampling, data is not visible for up to 420 seconds.
storage/stored_bytes ^GA Stored bytes
GAUGE, INT64, By bigquery_dataset	Number of bytes stored. Sampled every 1800 seconds. After sampling, data is not visible for up to 10800 seconds. table: Table name.
storage/table_count ^GA Table count
GAUGE, INT64, 1 bigquery_dataset	Number of tables. Sampled every 1800 seconds. After sampling, data is not visible for up to 10800 seconds.
storage/uploaded_bytes ^GA Uploaded bytes
DELTA, INT64, By bigquery_dataset	Uploaded bytes. Sampled every 60 seconds. After sampling, data is not visible for up to 21720 seconds. api: Api used to upload the data (batch import or streaming) table: Table name.
storage/uploaded_bytes_billed ^GA Uploaded bytes billed
DELTA, INT64, By bigquery_dataset	Uploaded bytes billed. Sampled every 60 seconds. After sampling, data is not visible for up to 21720 seconds. api: Api used to upload the data (batch import or streaming). Currently only streaming api is billed. table: Table name.
storage/uploaded_row_count ^GA Uploaded rows
DELTA, INT64, 1 bigquery_dataset	Uploaded rows. Sampled every 60 seconds. After sampling, data is not visible for up to 21720 seconds. api: Api used to upload the data (batch import or streaming) table: Table name.

bigquerybiengine

Metrics from the BigQuery service. Launch stages of these metrics: GA

The "metric type" strings in this table must be prefixed with bigquerybiengine.googleapis.com/. That prefix has been omitted from the entries in the table.

Metric type ^Launch stage Display name
Kind, Type, Unit Monitored resources	Description Labels
model/inflight_requests ^GA Inflight requests
GAUGE, INT64, 1 bigquery_biengine_model	Inflight requests to a BI model. Sampled every 60 seconds. After sampling, data is not visible for up to 180 seconds.
model/request_count ^GA Request count
DELTA, INT64, 1 bigquery_biengine_model	BI Engine request count. Sampled every 60 seconds. After sampling, data is not visible for up to 180 seconds. cache_miss: (BOOL) Indicates if request was not served from memory cache. Large number of requests with this field set to true is the sign that reservation size should be increased. . status: Completion status of the request. execution_mode: BIEngine or BigQuery. BigQuery means that request was served using BigQuery slots rather than using BI Engine reservation. .
model/request_latencies ^GA Request execution times
GAUGE, DISTRIBUTION, ms bigquery_biengine_model	Distribution of request execution times. Sampled every 60 seconds. After sampling, data is not visible for up to 180 seconds. cache_miss: (BOOL) Indicates if request was served from memory cache. Large number of requests with this field set to true is the sign that reservation size should be increased. . status: Completion status of the request. execution_mode: BIEngine or BigQuery. BigQuery means that request was served using BigQuery slots rather than using BI Engine reservation. .
reservation/total_bytes ^GA Reservation total bytes
GAUGE, INT64, By bigquery_project	Size of the BI engine reservation. Sampled every 60 seconds. After sampling, data is not visible for up to 180 seconds.
reservation/used_bytes ^GA Reservation used bytes
GAUGE, INT64, By bigquery_project	Utilization of the BI engine reservation. Sampled every 60 seconds. After sampling, data is not visible for up to 180 seconds.

bigtable

Metrics from Cloud Bigtable. Launch stages of these metrics: BETA EARLY_ACCESS GA

The "metric type" strings in this table must be prefixed with bigtable.googleapis.com/. That prefix has been omitted from the entries in the table.

Metric type ^Launch stage Display name
Kind, Type, Unit Monitored resources	Description Labels
backup/bytes_used ^BETA Backup storage used.
GAUGE, INT64, By bigtable_backup	Backup storage used in bytes. Sampled every 60 seconds. After sampling, data is not visible for up to 180 seconds. source_table: The source table of the backup. storage_type: Type of disk storage.
cluster/cpu_load ^GA CPU load
GAUGE, DOUBLE, 1 bigtable_cluster	CPU load of a cluster. Sampled every 60 seconds. After sampling, data is not visible for up to 180 seconds.
cluster/cpu_load_hottest_node ^GA CPU load (hottest node)
GAUGE, DOUBLE, 1 bigtable_cluster	CPU load of the busiest node in a cluster. Sampled every 60 seconds. After sampling, data is not visible for up to 180 seconds.
cluster/disk_load ^GA Disk load
GAUGE, DOUBLE, 1 bigtable_cluster	Utilization of HDD disks in a cluster. Sampled every 60 seconds. After sampling, data is not visible for up to 180 seconds.
cluster/node_count ^GA Nodes
GAUGE, INT64, 1 bigtable_cluster	Number of nodes in a cluster. Sampled every 60 seconds. After sampling, data is not visible for up to 180 seconds. storage_type: Storage type for the cluster.
cluster/storage_utilization ^GA Storage utilization
GAUGE, DOUBLE, 1 bigtable_cluster	Storage used as a fraction of total storage capacity. Sampled every 60 seconds. After sampling, data is not visible for up to 180 seconds. storage_type: Storage type for the cluster.
disk/bytes_used ^GA Data stored
GAUGE, INT64, By bigtable_cluster	Amount of compressed data for tables stored in a cluster. Sampled every 60 seconds. After sampling, data is not visible for up to 180 seconds. storage_type: Type of disk storage.
disk/per_node_storage_capacity ^EARLY_ACCESS Storage capacity per node
GAUGE, INT64, By bigtable_cluster	Capacity of compressed data for tables that can be stored per node in the cluster. Sampled every 60 seconds. After sampling, data is not visible for up to 180 seconds. storage_type: Storage type for the cluster.
disk/storage_capacity ^GA Storage capacity
GAUGE, INT64, By bigtable_cluster	Capacity of compressed data for tables that can be stored in a cluster. Sampled every 60 seconds. After sampling, data is not visible for up to 180 seconds. storage_type: Storage type for the cluster.
replication/latency ^GA Replication latencies
DELTA, DISTRIBUTION, ms bigtable_table	Distribution of replication request latencies for a table. Includes only requests that have been received by the destination cluster. Sampled every 60 seconds. After sampling, data is not visible for up to 240 seconds. source_cluster: Source cluster of the replicated data. source_zone: Source zone of the replicated data.
replication/max_delay ^GA Replication maximum delay
GAUGE, DOUBLE, s bigtable_table	Upper bound for replication delay between clusters of a table. Indicates the time frame during which latency information may not be accurate. Sampled every 60 seconds. After sampling, data is not visible for up to 240 seconds. source_cluster: Source cluster of the replicated data. source_zone: Source zone of the replicated data.
server/error_count ^GA Error count
DELTA, INT64, 1 bigtable_table	Number of server requests for a table that failed with an error. Sampled every 60 seconds. After sampling, data is not visible for up to 120 seconds. method: Cloud Bigtable API method. error_code: gRPC Error Code. app_profile: Cloud Bigtable application profile.
server/latencies ^GA Server Latencies
DELTA, DISTRIBUTION, ms bigtable_table	Distribution of server request latencies for a table, measured when calls reach Cloud Bigtable. Sampled every 60 seconds. After sampling, data is not visible for up to 120 seconds. method: Cloud Bigtable API method. app_profile: Cloud Bigtable application profile.
server/modified_rows_count ^GA Modified rows
DELTA, INT64, 1 bigtable_table	Number of rows modified by server requests for a table. Sampled every 60 seconds. After sampling, data is not visible for up to 120 seconds. method: Cloud Bigtable API method. app_profile: Cloud Bigtable application profile.
server/multi_cluster_failovers_count ^GA Multi-cluster failovers
DELTA, INT64, 1 bigtable_table	Number of failovers during multi-cluster requests. Sampled every 60 seconds. After sampling, data is not visible for up to 120 seconds. method: Cloud Bigtable API method. app_profile: Cloud Bigtable application profile.
server/received_bytes_count ^GA Received bytes
DELTA, INT64, By bigtable_table	Number of uncompressed bytes of request data received by servers for a table. Sampled every 60 seconds. After sampling, data is not visible for up to 120 seconds. method: Cloud Bigtable API method. app_profile: Cloud Bigtable application profile.
server/request_count ^GA Request count
DELTA, INT64, 1 bigtable_table	Number of server requests for a table. Sampled every 60 seconds. After sampling, data is not visible for up to 120 seconds. method: Cloud Bigtable API method. app_profile: Cloud Bigtable application profile.
server/returned_rows_count ^GA Returned rows
DELTA, INT64, 1 bigtable_table	Number of rows returned by server requests for a table. Sampled every 60 seconds. After sampling, data is not visible for up to 120 seconds. method: Cloud Bigtable API method. app_profile: Cloud Bigtable application profile.
server/sent_bytes_count ^GA Sent bytes
DELTA, INT64, By bigtable_table	Number of uncompressed bytes of response data sent by servers for a table. Sampled every 60 seconds. After sampling, data is not visible for up to 120 seconds. method: Cloud Bigtable API method. app_profile: Cloud Bigtable application profile.
table/bytes_used ^GA Data stored
GAUGE, INT64, By bigtable_table	Amount of compressed data stored in a table. Sampled every 60 seconds. After sampling, data is not visible for up to 120 seconds. storage_type: Type of disk storage.

cloudfunctions

Metrics from Cloud Functions. Launch stages of these metrics: BETA GA

The "metric type" strings in this table must be prefixed with cloudfunctions.googleapis.com/. That prefix has been omitted from the entries in the table.

Metric type ^Launch stage Display name
Kind, Type, Unit Monitored resources	Description Labels
function/active_instances ^BETA Active instances
GAUGE, INT64, 1 cloud_function	The number of active function instances. Sampled every 60 seconds. After sampling, data is not visible for up to 240 seconds.
function/execution_count ^GA Executions
DELTA, INT64, 1 cloud_function	Count of function executions broken down by status. Sampled every 60 seconds. After sampling, data is not visible for up to 240 seconds. memory: Memory assigned to function in MB. status: Execution status of the function: 'ok', 'timeout', 'error', 'crash', 'out of memory', 'out of quota', 'load error', 'load timeout', 'connection error', 'invalid header', 'request too large', 'system error', 'response error', 'invalid message'. trigger_type: Function's trigger type.
function/execution_times ^GA Execution times
DELTA, DISTRIBUTION, ns cloud_function	Distribution of functions execution times in nanoseconds. Sampled every 60 seconds. After sampling, data is not visible for up to 240 seconds. memory: Memory assigned to function in MB. status: Execution status of the function: 'ok', 'timeout', 'error', 'crash', 'out of memory', 'out of quota', 'load error', 'load timeout', 'connection error', 'invalid header', 'request too large', 'system error', 'response error', 'invalid message'. trigger_type: Function's trigger type.
function/network_egress ^GA Network egress
DELTA, INT64, By cloud_function	Outgoing network traffic of function, in bytes. Sampled every 60 seconds. After sampling, data is not visible for up to 240 seconds. memory: Memory assigned to function in MB. trigger_type: Function's trigger type.
function/user_memory_bytes ^GA Memory usage
DELTA, DISTRIBUTION, By cloud_function	Distribution of maximum function's memory usage during execution, in bytes. Sampled every 60 seconds. After sampling, data is not visible for up to 240 seconds. memory: Memory assigned to function in MB. trigger_type: Function's trigger type.

cloudiot

Metrics from IoT Core. Launch stages of these metrics: GA

The "metric type" strings in this table must be prefixed with cloudiot.googleapis.com/. That prefix has been omitted from the entries in the table.

Metric type ^Launch stage Display name
Kind, Type, Unit Monitored resources	Description Labels
device/active_devices ^GA Active devices
GAUGE, INT64, 1 cloudiot_device_registry	Count of the number of devices that have recently sent data to Cloud IoT Core. Sampled every 60 seconds. After sampling, data is not visible for up to 180 seconds. protocol: Device protocol (eg. `http`, `mqtt`).
device/billing_bytes_count ^GA Billable bytes transferred
GAUGE, INT64, By cloudiot_device_registry	Count of the number of billable bytes transferred by devices. Sampled every 60 seconds. After sampling, data is not visible for up to 180 seconds. protocol: Device protocol (eg. `http`, `mqtt`). direction: Direction of data traffic (`ingress` or `egress`).
device/error_count ^GA Errors communicating with devices
DELTA, INT64, 1 cloudiot_device_registry	Delta count of errors communicating with devices, grouped by error type. Sampled every 60 seconds. After sampling, data is not visible for up to 180 seconds. error: Type of error. For example, 'pubsub_publish', or 'incoming_buffer_full'. protocol: Device protocol (eg. 'mqtt', 'http').
device/operation_count ^GA Operations
DELTA, INT64, 1 cloudiot_device_registry	Delta count of operations performed, grouped by operation type. Sampled every 60 seconds. After sampling, data is not visible for up to 180 seconds. operation_type: Type of operation performed.
device/received_bytes_count ^GA Bytes received by devices
DELTA, INT64, By cloudiot_device_registry	Delta count of number of bytes received from devices. Sampled every 60 seconds. After sampling, data is not visible for up to 180 seconds.
device/sent_bytes_count ^GA Bytes sent to devices
DELTA, INT64, By cloudiot_device_registry	Delta count of number of bytes sent by devices. Sampled every 60 seconds. After sampling, data is not visible for up to 180 seconds.

cloudsql

Metrics from Cloud SQL. Launch stages of these metrics: BETA GA

The "metric type" strings in this table must be prefixed with cloudsql.googleapis.com/. That prefix has been omitted from the entries in the table.

Metric type ^Launch stage Display name
Kind, Type, Unit Monitored resources	Description Labels
database/auto_failover_request_count ^GA Auto-failover Requests
DELTA, INT64, 1 cloudsql_database	Delta of number of instance auto-failover requests. Sampled every 60 seconds. After sampling, data is not visible for up to 210 seconds.
database/available_for_failover ^GA Available for failover
GAUGE, INT64, 1 cloudsql_database	This is > 0 if the failover operation is available on the instance. Sampled every 60 seconds. After sampling, data is not visible for up to 210 seconds.
database/cpu/reserved_cores ^GA CPU reserved cores
GAUGE, DOUBLE, 1 cloudsql_database	Number of cores reserved for the database. Sampled every 60 seconds. After sampling, data is not visible for up to 210 seconds.
database/cpu/usage_time ^GA CPU usage
DELTA, DOUBLE, s{CPU} cloudsql_database	Cumulative CPU usage time in seconds. Sampled every 60 seconds. After sampling, data is not visible for up to 210 seconds.
database/cpu/utilization ^GA CPU utilization
GAUGE, DOUBLE, 10^2.% cloudsql_database	Current CPU utilization represented as a percentage of the reserved CPU that is currently in use. Values are typically numbers between 0.0 and 1.0 (but might exceed 1.0). Charts display the values as a percentage between 0% and 100% (or more). Sampled every 60 seconds. After sampling, data is not visible for up to 150 seconds.
database/disk/bytes_used ^GA Bytes used
GAUGE, INT64, By cloudsql_database	Data utilization in bytes. Sampled every 60 seconds. After sampling, data is not visible for up to 210 seconds.
database/disk/quota ^GA Disk quota
GAUGE, INT64, By cloudsql_database	Maximum data disk size in bytes. Sampled every 60 seconds. After sampling, data is not visible for up to 210 seconds.
database/disk/read_ops_count ^GA Disk read IO
DELTA, INT64, 1 cloudsql_database	Delta count of data disk read IO operations. Sampled every 60 seconds. After sampling, data is not visible for up to 210 seconds.
database/disk/utilization ^GA Disk utilization
GAUGE, DOUBLE, 1 cloudsql_database	The fraction of the disk quota that is currently in use. Sampled every 60 seconds. After sampling, data is not visible for up to 150 seconds.
database/disk/write_ops_count ^GA Disk write IO
DELTA, INT64, 1 cloudsql_database	Delta count of data disk write IO operations. Sampled every 60 seconds. After sampling, data is not visible for up to 210 seconds.
database/instance_state ^GA Instance state
GAUGE, BOOL,  cloudsql_database	The current serving state of the Cloud SQL instance. This can be one of the following: 1 (RUNNING): The instance is *expected* to be running. If an instance experiences unplanned (non-maintenance) downtime, the state will still be RUNNING, but the database/up metric will report 0. 2 (SUSPENDED): The instance is not available, for example due to problems with billing. 3 (RUNNABLE): The instance has been stopped by owner. It is not currently running, but it's ready to be restarted. 4 (PENDING_CREATE): The instance is being created. 5 (MAINTENANCE): The instance is down for maintenance. 6 (FAILED): The instance creation failed or an operation left the instance in an unknown bad state. 7 (UNKNOWN_STATE): The state of the instance is unknown. Sampled every 60 seconds. After sampling, data is not visible for up to 210 seconds. state: The state of the instance.
database/memory/quota ^GA Memory quota
GAUGE, INT64, By cloudsql_database	Maximum RAM size in bytes. Sampled every 60 seconds. After sampling, data is not visible for up to 210 seconds.
database/memory/total_usage ^GA Total memory usage
GAUGE, INT64, By cloudsql_database	Total RAM usage in bytes including buffer cache. Sampled every 60 seconds. After sampling, data is not visible for up to 210 seconds.
database/memory/usage ^GA Memory usage
GAUGE, INT64, By cloudsql_database	RAM usage in bytes. Sampled every 60 seconds. After sampling, data is not visible for up to 210 seconds.
database/memory/utilization ^GA Memory utilization
GAUGE, DOUBLE, 1 cloudsql_database	The fraction of the memory quota that is currently in use. Sampled every 60 seconds. After sampling, data is not visible for up to 150 seconds.
database/mysql/innodb_buffer_pool_pages_dirty ^GA InnoDB dirty pages
GAUGE, INT64, 1 cloudsql_database	Number of unflushed pages in the InnoDB buffer pool. Sampled every 60 seconds. After sampling, data is not visible for up to 210 seconds.
database/mysql/innodb_buffer_pool_pages_free ^GA InnoDB free pages
GAUGE, INT64, 1 cloudsql_database	Number of unused pages in the InnoDB buffer pool. Sampled every 60 seconds. After sampling, data is not visible for up to 210 seconds.
database/mysql/innodb_buffer_pool_pages_total ^GA InnoDB total pages
GAUGE, INT64, 1 cloudsql_database	Total number of pages in the InnoDB buffer pool. Sampled every 60 seconds. After sampling, data is not visible for up to 210 seconds.
database/mysql/innodb_data_fsyncs ^GA InnoDB fsync calls
DELTA, INT64, 1 cloudsql_database	Delta count of InnoDB fsync() calls. Sampled every 60 seconds. After sampling, data is not visible for up to 210 seconds.
database/mysql/innodb_os_log_fsyncs ^GA InnoDB log fsync calls
DELTA, INT64, 1 cloudsql_database	Delta count of InnoDB fsync() calls to the log file. Sampled every 60 seconds. After sampling, data is not visible for up to 210 seconds.
database/mysql/innodb_pages_read ^GA InnoDB pages read
DELTA, INT64, 1 cloudsql_database	Delta count of InnoDB pages read. Sampled every 60 seconds. After sampling, data is not visible for up to 210 seconds.
database/mysql/innodb_pages_written ^GA InnoDB pages written
DELTA, INT64, 1 cloudsql_database	Delta count of InnoDB pages written. Sampled every 60 seconds. After sampling, data is not visible for up to 210 seconds.
database/mysql/queries ^GA Queries
DELTA, INT64, 1 cloudsql_database	Delta count of statements executed by the server. Sampled every 60 seconds. After sampling, data is not visible for up to 210 seconds.
database/mysql/questions ^GA Questions
DELTA, INT64, 1 cloudsql_database	Delta count of statements executed by the server sent by the client. Sampled every 60 seconds. After sampling, data is not visible for up to 210 seconds.
database/mysql/received_bytes_count ^GA Network bytes received by MySQL
DELTA, INT64, By cloudsql_database	Delta count of bytes received by MySQL process. Sampled every 60 seconds. After sampling, data is not visible for up to 210 seconds.
database/mysql/replication/available_for_failover ^DEPRECATED Available for failover (Deprecated)
GAUGE, INT64, 1 cloudsql_database	This is > 0 if the failover operation is available on the master instance.master. The metric is deprecated. Please use cloudsql.googleapis.com/database/available_for_failover instead. Sampled every 60 seconds. After sampling, data is not visible for up to 210 seconds.
database/mysql/replication/last_io_errno ^GA Last I/O thread error number.
GAUGE, INT64, 1 cloudsql_database	The error number of the most recent error that caused the I/O thread to stop. Sampled every 60 seconds. After sampling, data is not visible for up to 210 seconds.
database/mysql/replication/last_sql_errno ^GA Last SQL thread error number.
GAUGE, INT64, 1 cloudsql_database	The error number of the most recent error that caused the SQL thread to stop. Sampled every 60 seconds. After sampling, data is not visible for up to 210 seconds.
database/mysql/replication/seconds_behind_master ^GA Replica lag
GAUGE, INT64, s cloudsql_database	Number of seconds the read replica is behind its master (approximation). Sampled every 60 seconds. After sampling, data is not visible for up to 210 seconds.
database/mysql/replication/slave_io_running ^GA Slave I/O thread running
GAUGE, STRING,  cloudsql_database	Indicates whether the I/O thread for reading the master's binary log is running. Possible values are Yes, No and Connecting. Sampled every 60 seconds. After sampling, data is not visible for up to 60 seconds.
database/mysql/replication/slave_io_running_state ^GA Slave I/O thread running state
GAUGE, BOOL,  cloudsql_database	Indicates whether the I/O thread for reading the master's binary log is running. Possible values are Yes, No and Connecting, and the values are exposed through the 'state' field. Sampled every 60 seconds. After sampling, data is not visible for up to 210 seconds. state: The slave_io_running state of the slave.
database/mysql/replication/slave_sql_running ^GA Slave SQL thread running
GAUGE, STRING,  cloudsql_database	Indicates whether the SQL thread for executing events in the relay log is running. Sampled every 60 seconds. After sampling, data is not visible for up to 60 seconds.
database/mysql/replication/slave_sql_running_state ^GA Slave SQL thread running state
GAUGE, BOOL,  cloudsql_database	Indicates whether the SQL thread for executing events in the relay log is running. Possible values are Yes / No, and the values are exposed through the 'state' field. Sampled every 60 seconds. After sampling, data is not visible for up to 210 seconds. state: The slave_sql_running state of the slave.
database/mysql/sent_bytes_count ^GA Network bytes sent by MySQL
DELTA, INT64, By cloudsql_database	Delta count of bytes sent by MySQL process. Sampled every 60 seconds. After sampling, data is not visible for up to 210 seconds.
database/network/connections ^GA Cloud SQL Connections
GAUGE, INT64, 1 cloudsql_database	Number of connections to databases on the Cloud SQL instance. Sampled every 60 seconds. After sampling, data is not visible for up to 210 seconds.
database/network/received_bytes_count ^GA Received bytes
DELTA, INT64, By cloudsql_database	Delta count of bytes received through the network. Sampled every 60 seconds. After sampling, data is not visible for up to 210 seconds.
database/network/sent_bytes_count ^GA Sent bytes
DELTA, INT64, By cloudsql_database	Delta count of bytes sent through the network. Sampled every 60 seconds. After sampling, data is not visible for up to 210 seconds. destination: Where the data is being sent.
database/postgresql/insights/aggregate/execution_time ^BETA Execution time
CUMULATIVE, INT64, us{CPU} cloudsql_instance_database	Accumulated query execution time per user per database. This is the sum of cpu time, io wait time, lock wait time, process context switch, and scheduling for all the processes involved in the query execution. user: The user of the database. client_addr: The IP address of the database client.
database/postgresql/insights/aggregate/io_time ^BETA IO time
CUMULATIVE, INT64, us cloudsql_instance_database	Accumulated IO time per user per database. user: The user of the database. client_addr: The IP address of the database client. io_type: The IO type (read/write).
database/postgresql/insights/aggregate/latencies ^BETA Latency
CUMULATIVE, DISTRIBUTION, us cloudsql_instance_database	Query latency distribution per user per database. user: The user of the database. client_addr: The IP address of the database client.
database/postgresql/insights/aggregate/lock_time ^BETA Aggregated lock time
CUMULATIVE, INT64, us cloudsql_instance_database	Accumulated lock wait time per user per database. user: The user of the database. client_addr: The IP address of the database client. lock_type: The lock type (lw/hw/buffer_pin).
database/postgresql/insights/aggregate/row_count ^BETA Affected rows
CUMULATIVE, INT64, 1 cloudsql_instance_database	Total number of retrieved or affected rows. user: The user of the database. client_addr: The IP address of the database client.
database/postgresql/insights/aggregate/shared_blk_access_count ^BETA Shared blocks cache access.
CUMULATIVE, INT64, 1 cloudsql_instance_database	Shared blocks (regular tables & indexed) accessed by statement execution. user: The user of the database. client_addr: The IP address of the database client. access_type: The block access type (hit/miss).
database/postgresql/insights/perquery/execution_time ^BETA Per query execution times
CUMULATIVE, INT64, us{CPU} cloudsql_instance_database	Accumulated execution times per user per database per query.This is the sum of cpu time, io wait time, lock wait time, process context switch, and scheduling for all the processes involved in the query execution. querystring: The normalized query. user: The user of the database. client_addr: The IP address of the database client. query_hash: A unique hash of the query string.
database/postgresql/insights/perquery/io_time ^BETA Per query IO time
CUMULATIVE, INT64, us cloudsql_instance_database	Accumulated IO time per user per database per query. querystring: The normalized query. user: The user of the database. client_addr: The IP address of the database client. io_type: The IO type (read/write). query_hash: A unique hash of the query string.
database/postgresql/insights/perquery/latencies ^BETA Per query latency
CUMULATIVE, DISTRIBUTION, us cloudsql_instance_database	Query latency distribution per user per database per query. querystring: The normalized query. user: The user of the database. client_addr: The IP address of the database client. query_hash: A unique hash of the query string.
database/postgresql/insights/perquery/lock_time ^BETA Per query lock time
CUMULATIVE, INT64, us cloudsql_instance_database	Accumulated lock wait time per user per database per query. querystring: The normalized query. user: The user of the database. client_addr: The IP address of the database client. lock_type: The lock type (lw/hw/buffer_pin). query_hash: A unique hash of the query string.
database/postgresql/insights/perquery/row_count ^BETA Per query affected rows
CUMULATIVE, INT64, 1 cloudsql_instance_database	Total number of retrieved or affected rows. querystring: The normalized query. user: The user of the database. client_addr: The IP address of the database client. query_hash: A unique hash of the query string.
database/postgresql/insights/perquery/shared_blk_access_count ^BETA Per query Shared blocks cache access
CUMULATIVE, INT64, 1 cloudsql_instance_database	Shared blocks (regular tables & indexed) accesssed by statement execution. querystring: The normalized query. user: The user of the database. client_addr: The IP address of the database client. access_type: The block access type (hit/miss). query_hash: A unique hash of the query string.
database/postgresql/insights/pertag/execution_time ^BETA Per tag execution time
CUMULATIVE, INT64, us{CPU} cloudsql_instance_database	Accumulated execution times per user per database per tag.This is the sum of cpu time, io wait time, lock wait time, process context switch, and scheduling for all the processes involved in the query execution. user: The user of the database. client_addr: The IP address of the database client. action: The name of the action. application: The name of the application. controller: The name of the controller. db_driver: The name of the db_driver. framework: The name of the framework. route: The name of the route. tag_hash: A unique hash of the tag string.
database/postgresql/insights/pertag/io_time ^BETA Per tag IO time
CUMULATIVE, INT64, us cloudsql_instance_database	Accumulated IO write time per user per database per tag. user: The user of the database. client_addr: The IP address of the database client. action: The name of the action. application: The name of the application. controller: The name of the controller. db_driver: The name of the db_driver. framework: The name of the framework. route: The name of the route. io_type: The IO type (read/write). tag_hash: A unique hash of the tag string.
database/postgresql/insights/pertag/latencies ^BETA Per tag latency
CUMULATIVE, DISTRIBUTION, us cloudsql_instance_database	Query latency distribution per user per database per tag. user: The user of the database. client_addr: The IP address of the database client. action: The name of the action. application: The name of the application. controller: The name of the controller. db_driver: The name of the db_driver. framework: The name of the framework. route: The name of the route. tag_hash: A unique hash of the tag string.
database/postgresql/insights/pertag/lock_time ^BETA Per tag lock time
CUMULATIVE, INT64, us cloudsql_instance_database	Accumulated lock wait time per user per database per tag. user: The user of the database. client_addr: The IP address of the database client. action: The name of the action. application: The name of the application. controller: The name of the controller. db_driver: The name of the db_driver. framework: The name of the framework. route: The name of the route. lock_type: The lock type (lw/hw/buffer_pin). tag_hash: A unique hash of the tag string.
database/postgresql/insights/pertag/row_count ^BETA Per tag affected rows
CUMULATIVE, INT64, 1 cloudsql_instance_database	Total number of retrieved or affected rows. user: The user of the database. client_addr: The IP address of the database client. action: The name of the action. application: The name of the application. controller: The name of the controller. db_driver: The name of the db_driver. framework: The name of the framework. route: The name of the route. tag_hash: A unique hash of the tag string.
database/postgresql/insights/pertag/shared_blk_access_count ^BETA Per tag shared blocks cache accessed
CUMULATIVE, INT64, 1 cloudsql_instance_database	Shared blocks (regular tables & indexed) accessed by statement execution. user: The user of the database. client_addr: The IP address of the database client. action: The name of the action. application: The name of the application. controller: The name of the controller. db_driver: The name of the db_driver. framework: The name of the framework. route: The name of the route. access_type: The block access type (hit/miss). tag_hash: A unique hash of the tag string.
database/postgresql/num_backends ^GA PostgreSQL Connections
GAUGE, INT64, 1 cloudsql_database	Number of connections to the Cloud SQL PostgreSQL instance. Sampled every 60 seconds. After sampling, data is not visible for up to 210 seconds. database: Name of the database.
database/postgresql/replication/replica_byte_lag ^GA Lag bytes
GAUGE, INT64, By cloudsql_database	Replication lag in bytes. Reported from the master per replica. Sampled every 60 seconds. After sampling, data is not visible for up to 210 seconds. replica_name: Name of the replica. replica_lag_type: Type of lag. This can be either replay_location, flush_location, write_location or sent_location.
database/postgresql/transaction_count ^GA Number of transactions
DELTA, INT64, 1 cloudsql_database	Delta count of number of transactions. Sampled every 60 seconds. After sampling, data is not visible for up to 210 seconds. database: Name of the database. transaction_type: transaction_type can be commit or rollback.
database/replication/replica_lag ^GA Replica lag
GAUGE, DOUBLE, s cloudsql_database	Number of seconds the read replica is behind its master (approximation). Sampled every 60 seconds. After sampling, data is not visible for up to 210 seconds.
database/replication/state ^BETA Replication state
GAUGE, BOOL,  cloudsql_database	The current serving state of replication. Sampled every 60 seconds. After sampling, data is not visible for up to 210 seconds. state: State of replication:"Running" - Replication is active and running;"Stopped" - Replication is inactive and stopped;"Error" - There is an error happening and replication is stopped.
database/state ^GA State
GAUGE, STRING,  cloudsql_database	The current serving state of the Cloud SQL instance. This can be one of the following: 1 (RUNNING): The instance is *expected* to be running. If an instance experiences unplanned (non-maintenance) downtime, the state will still be RUNNING, but the database/up metric will report 0. 2 (SUSPENDED): The instance is not available, for example due to problems with billing. 3 (RUNNABLE): The instance has been stopped by owner. It is not currently running, but it's ready to be restarted. 4 (PENDING_CREATE): The instance is being created. 5 (MAINTENANCE): The instance is down for maintenance. 6 (FAILED): The instance creation failed or an operation left the instance in an unknown bad state. 7 (UNKNOWN_STATE): The state of the instance is unknown. Sampled every 60 seconds. After sampling, data is not visible for up to 210 seconds.
database/up ^GA Server up
GAUGE, INT64, 1 cloudsql_database	Indicates if the server is up or not. On-demand instances are spun down if no connections are made for a sufficient amount of time. Sampled every 60 seconds. After sampling, data is not visible for up to 210 seconds.
database/uptime ^GA Uptime
DELTA, INT64, s cloudsql_database	Delta count of the time in seconds the instance has been running. Sampled every 60 seconds. After sampling, data is not visible for up to 210 seconds.

cloudtasks

Metrics from Cloud Tasks (formerly App Engine Task Queue).

Note: This service was previously called “App Engine Task Queue”, and some of the metrics were published with the prefix taskqueue.googleapis.com.

Launch stages of these metrics: BETA GA

The "metric type" strings in this table must be prefixed with cloudtasks.googleapis.com/. That prefix has been omitted from the entries in the table.

Metric type ^Launch stage Display name
Kind, Type, Unit Monitored resources	Description Labels
api/request_count ^GA API requests
DELTA, INT64, 1 cloud_tasks_queue	Count of Cloud Tasks API calls. Sampled every 60 seconds. After sampling, data is not visible for up to 240 seconds. api_method: The API method called (e.g. CreateTask). response_code: Canonical response code as string (e.g. 'ok').
queue/depth ^BETA Queue depth
GAUGE, INT64, 1 cloud_tasks_queue	The number of tasks in the queue. Sampled every 60 seconds. After sampling, data is not visible for up to 120 seconds.
queue/task_attempt_count ^GA Task attempt count
DELTA, INT64, 1 cloud_tasks_queue	Count of task attempts broken down by response code. Sampled every 60 seconds. After sampling, data is not visible for up to 180 seconds. response_code: Canonical response code as string (e.g. 'ok').
queue/task_attempt_delays ^GA Task attempt delays
DELTA, DISTRIBUTION, ms cloud_tasks_queue	Delay between each scheduled attempt time and actual attempt time. Sampled every 60 seconds. After sampling, data is not visible for up to 120 seconds.

cloudtrace

Metrics from Cloud Trace. Launch stages of these metrics: BETA GA

The "metric type" strings in this table must be prefixed with cloudtrace.googleapis.com/. That prefix has been omitted from the entries in the table.

Metric type ^Launch stage Display name
Kind, Type, Unit Monitored resources	Description Labels
bigquery_export/exported_span_count ^BETA Spans Exported to BigQuery
DELTA, INT64, 1 cloudtrace.googleapis.com/CloudtraceProject	Counts per status of attempts to export spans to BigQuery via a TraceSink. Note that spans will be counted for each configured TraceSink that they match. status: The status of an attempt to export spans to BigQuery via a TraceSink.
billing/monthly_spans_ingested ^GA Monthly trace spans ingested
GAUGE, INT64, 1 global	Month-to-date trace spans ingested. Sampled every 1800 seconds. After sampling, data is not visible for up to 6000 seconds. service: Service associated with the ingested span. chargeable: (BOOL) Whether the span is chargeable.
billing/spans_ingested ^GA Trace spans ingested
DELTA, INT64, 1 global	Trace spans ingested. Sampled every 60 seconds. After sampling, data is not visible for up to 300 seconds. service: Service associated with the ingested span. chargeable: (BOOL) Whether the span is chargeable.

composer

Metrics from Cloud Composer. Launch stages of these metrics: BETA

The "metric type" strings in this table must be prefixed with composer.googleapis.com/. That prefix has been omitted from the entries in the table.

Metric type ^Launch stage Display name
Kind, Type, Unit Monitored resources	Description Labels
environment/api/request_count ^BETA API Requests
DELTA, INT64, 1 cloud_composer_environment	Number of Composer API requests seen so far. Sampled every 60 seconds. After sampling, data is not visible for up to 180 seconds. status: The status code of the API call. api_version: The version of the Composer API, e.g., v1beta1. api_method: The name of the Composer API call, e.g., CreateEnvironment.
environment/api/request_latencies ^BETA API Latency
DELTA, DISTRIBUTION, ms cloud_composer_environment	Distribution of Composer API call latencies. Sampled every 60 seconds. After sampling, data is not visible for up to 180 seconds. status: The status code of the API call. api_version: The version of the Composer API, e.g., v1beta1. api_method: The name of the Composer API call, e.g., CreateEnvironment.
environment/dag_processing/parse_error_count ^BETA Parse Error Count
DELTA, INT64, 1 cloud_composer_environment	Number of errors raised during parsing DAG files. Sampled every 60 seconds. image_version: Version of Composer image.
environment/dag_processing/processes ^BETA DAG parsing processes
GAUGE, INT64, 1 cloud_composer_environment	Number of currently running DAG parsing processes. Sampled every 60 seconds. image_version: Version of Composer image.
environment/dag_processing/processor_timeout_count ^BETA Processors Timeout Count
DELTA, INT64, 1 cloud_composer_environment	Number of file processors terminated due to processing timeout. Sampled every 60 seconds. image_version: Version of Composer image.
environment/dag_processing/total_parse_time ^BETA Total Parse Time
GAUGE, DOUBLE, s cloud_composer_environment	Number of seconds taken to scan and import all DAG files once. Sampled every 60 seconds. image_version: Version of Composer image.
environment/dagbag_size ^BETA Dag Bag Size
GAUGE, INT64, 1 cloud_composer_environment	The current dag bag size. Sampled every 60 seconds. image_version: Version of Composer image.
environment/database/cpu/reserved_cores ^BETA Database CPU reserved cores
GAUGE, DOUBLE, 1 cloud_composer_environment	Number of cores reserved for the database instance. Sampled every 60 seconds. image_version: Version of Composer image.
environment/database/cpu/usage_time ^BETA Database CPU usage time
DELTA, DOUBLE, s cloud_composer_environment	CPU usage time of the database instance. Sampled every 60 seconds. image_version: Version of Composer image.
environment/database/cpu/utilization ^BETA Database CPU utilization
GAUGE, DOUBLE, 1 cloud_composer_environment	CPU utilization ratio (from 0.0 to 1.0) of the database instance. Sampled every 60 seconds. image_version: Version of Composer image.
environment/database/disk/bytes_used ^BETA Database disk usage
GAUGE, INT64, By cloud_composer_environment	Used disk space in bytes on the database instance. Sampled every 60 seconds. image_version: Version of Composer image.
environment/database/disk/quota ^BETA Database disk quota
GAUGE, INT64, By cloud_composer_environment	Maximum data disk size in bytes of the database instance. Sampled every 60 seconds. image_version: Version of Composer image.
environment/database/disk/utilization ^BETA Database disk utilization
GAUGE, DOUBLE, 1 cloud_composer_environment	Disk quota usage ratio (from 0.0 to 1.0) of the database instance. Sampled every 60 seconds. image_version: Version of Composer image.
environment/database/memory/bytes_used ^BETA Database memory usage
GAUGE, INT64, By cloud_composer_environment	Memory usage of the database instance in bytes. Sampled every 60 seconds. image_version: Version of Composer image.
environment/database/memory/quota ^BETA Database memory quota
GAUGE, INT64, By cloud_composer_environment	Maximum RAM size in bytes of the database instance. Sampled every 60 seconds. image_version: Version of Composer image.
environment/database/memory/utilization ^BETA Database memory utilization
GAUGE, DOUBLE, 1 cloud_composer_environment	Memory utilization ratio (from 0.0 to 1.0) of the database instance. Sampled every 60 seconds. image_version: Version of Composer image.
environment/database_health ^BETA Database Healthy
GAUGE, BOOL,  cloud_composer_environment	Healthiness of Composer airflow database. Sampled every 60 seconds. image_version: Version of Composer image. environment_type: Environment type, e.g. private_ip.
environment/executor/open_slots ^BETA Executor Open Slots
GAUGE, INT64, 1 cloud_composer_environment	Number of open slots on executor. Sampled every 60 seconds. image_version: Version of Composer image.
environment/executor/running_tasks ^BETA Executor Running Tasks
GAUGE, INT64, 1 cloud_composer_environment	Number of running tasks on executor. Sampled every 60 seconds. image_version: Version of Composer image.
environment/finished_task_instance_count ^BETA Task Instance Count
DELTA, INT64, 1 cloud_composer_environment	Overall task instances. Sampled every 60 seconds. state: State of task instance. image_version: Version of Composer image.
environment/healthy ^BETA Healthy
GAUGE, BOOL,  cloud_composer_environment	Healthiness of Composer environment. Sampled every 300 seconds. image_version: Version of Composer image.
environment/num_celery_workers ^BETA Celery Workers
GAUGE, INT64, 1 cloud_composer_environment	Number of Celery workers. Sampled every 60 seconds. image_version: Version of Composer image.
environment/scheduler_heartbeat_count ^BETA Scheduler Heartbeats
DELTA, INT64, 1 cloud_composer_environment	Scheduler heartbeats. Sampled every 60 seconds. image_version: Version of Composer image.
environment/task_queue_length ^BETA Task Queue Length
GAUGE, INT64, 1 cloud_composer_environment	Number of tasks in queue. Sampled every 60 seconds. image_version: Version of Composer image. queue_name: Name of task queue.
environment/web_server/cpu/reserved_cores ^BETA Web server CPU reserved cores
GAUGE, DOUBLE, 1 cloud_composer_environment	Number of cores reserved for the web server instance. Sampled every 60 seconds. image_version: Version of Composer image. instance_id: Numeric ID of GCE instance running the web server.
environment/web_server/cpu/usage_time ^BETA Web server CPU usage time
DELTA, DOUBLE, s cloud_composer_environment	CPU usage time of the web server instance. Sampled every 60 seconds. image_version: Version of Composer image. instance_id: Numeric ID of GCE instance running the web server.
environment/web_server/memory/bytes_used ^BETA Web server memory usage
GAUGE, INT64, By cloud_composer_environment	Memory usage of the web server instance in bytes. Sampled every 60 seconds. image_version: Version of Composer image. instance_id: Numeric ID of GCE instance running the web server.
environment/web_server/memory/quota ^BETA Web server memory quota
GAUGE, INT64, By cloud_composer_environment	Maximum RAM size in bytes of the web server instance. Sampled every 60 seconds. image_version: Version of Composer image. instance_id: Numeric ID of GCE instance running the web server.
environment/worker/pod_eviction_count ^BETA Worker Pod Eviction Count
DELTA, INT64, 1 cloud_composer_environment	The number of Airflow worker pods evictions. Sampled every 60 seconds. image_version: Version of Composer image.
environment/worker/scale_factor_target ^BETA Workers scale factor target
GAUGE, DOUBLE, 1 cloud_composer_environment	Scale factor for Airflow workers count. Sampled every 60 seconds. image_version: Version of Composer image. queue_name: Name of the Airflow queue. See https://airflow.apache.org/docs/stable/concepts.html#queues.
environment/zombie_task_killed_count ^BETA Zombie Tasks Killed
DELTA, INT64, 1 cloud_composer_environment	Number of zombie tasks killed. Sampled every 60 seconds. image_version: Version of Composer image.
workflow/run_count ^BETA Workflow Runs
DELTA, INT64, 1 cloud_composer_workflow	Number of workflow runs completed so far. Sampled every 60 seconds. state: State of workflow run. image_version: Version of Composer image.
workflow/run_duration ^BETA Workflow Run Duration
GAUGE, DOUBLE, s cloud_composer_workflow	Duration of workflow run completion. Sampled every 60 seconds. state: State of workflow run. image_version: Version of Composer image.
workflow/task/run_count ^BETA Tasks
DELTA, INT64, 1 cloud_composer_workflow	Number of workflow tasks completed so far. Sampled every 60 seconds. state: State of task. image_version: Version of Composer image. task_name: Name of task. operator: Operator of task.
workflow/task/run_duration ^BETA Task Duration
GAUGE, DOUBLE, s cloud_composer_workflow	Duration of task completion. Sampled every 60 seconds. state: State of task. image_version: Version of Composer image. task_name: Name of task. operator: Operator of task.

compute

Metrics from Compute Engine. These metrics do not require installation of the Monitoring or Logging agent, but to collect the compute.googleapis.com/guest metrics, you must enable the Container-Optimized OS Health Monitoring feature; for more information, see Container-Optimized OS. Launch stages of these metrics: ALPHA BETA GA

The "metric type" strings in this table must be prefixed with compute.googleapis.com/. That prefix has been omitted from the entries in the table.

Metric type ^Launch stage Display name
Kind, Type, Unit Monitored resources	Description Labels
firewall/dropped_bytes_count ^GA Dropped bytes
DELTA, INT64, By gce_instance	Delta count of incoming bytes dropped by the firewall. Sampled every 60 seconds. After sampling, data is not visible for up to 240 seconds. instance_name: The name of the VM instance.
firewall/dropped_packets_count ^GA Dropped packets
DELTA, INT64, 1 gce_instance	Delta count of incoming packets dropped by the firewall. Sampled every 60 seconds. After sampling, data is not visible for up to 240 seconds. instance_name: The name of the VM instance.
guest/cpu/load_15m ^ALPHA CPU load average (15m)
GAUGE, DOUBLE, 1 gce_instance	CPU Load (15 minute intervals). Container-Optimized OS only. instance_name: The name of the VM instance.
guest/cpu/load_1m ^ALPHA CPU load average (1m)
GAUGE, DOUBLE, 1 gce_instance	CPU Load (1 minute intervals). Container-Optimized OS only. instance_name: The name of the VM instance.
guest/cpu/load_5m ^ALPHA CPU load average (5m)
GAUGE, DOUBLE, 1 gce_instance	CPU Load (5 minute intervals). Container-Optimized OS only. instance_name: The name of the VM instance.
guest/cpu/runnable_task_count ^BETA Runnable task count.
GAUGE, DOUBLE, 1 gce_instance	The average number of runnable tasks in the run-queue. Container-Optimized OS only. instance_name: The name of the VM instance.
guest/cpu/usage_time ^BETA CPU usage
CUMULATIVE, DOUBLE, s{CPU} gce_instance	CPU usage, in seconds. Container-Optimized OS only. instance_name: The name of the VM instance. state: One of the following: idle, interrupt, nice, softirq, steal, system, user, wait.
guest/disk/bytes_used ^BETA Disk usage
GAUGE, INT64, By gce_instance	Number of bytes used on disk for file systems. Container-Optimized OS only. instance_name: The name of the VM instance. mount_option: Mount options on the disk device. device_name: The name of the disk device. state: Type of usage (free, used, reserved) fs_type: File system type of the disk device.
guest/disk/io_time ^GA IO Time
CUMULATIVE, INT64, ms gce_instance	The cumulative time spent on the I/O operations that are in progress; that is, the actual time in queue and when disks were busy. Requests issued in parallel are counted as a single one. Container-Optimized OS only. Sampled every 60 seconds. instance_name: The name of the VM instance. device_name: The name of the disk device.
guest/disk/merged_operation_count ^BETA Merged disk operations
CUMULATIVE, INT64, 1 gce_instance	Merged disk operations count. Disk operations which are adjacent to each other may be merged by the kernel for efficiency. Container-Optimized OS only. instance_name: The name of the VM instance. device_name: The name of the disk device. direction: Direction (read, write).
guest/disk/operation_bytes_count ^BETA Disk bytes transferred
CUMULATIVE, INT64, By gce_instance	Bytes transferred in disk operations. Container-Optimized OS only. instance_name: The name of the VM instance. device_name: The name of the disk device. direction: Direction (read, write).
guest/disk/operation_count ^BETA Disk operations
CUMULATIVE, INT64, 1 gce_instance	Disk operations count. Container-Optimized OS only. instance_name: The name of the VM instance. device_name: The name of the disk device. direction: Direction (read, write).
guest/disk/operation_time ^BETA Disk operation time
CUMULATIVE, INT64, ms gce_instance	Amount of time spent on the disk operations, by direction. This metric only includes time spent on completed operations. Container-Optimized OS only. instance_name: The name of the VM instance. device_name: The name of the disk device. direction: Direction (read, write).
guest/disk/queue_length ^GA Queue Length
GAUGE, DOUBLE, 1 gce_instance	The queue length on the disk averaged over the last 60 seconds. Container-Optimized OS only. Sampled every 60 seconds. instance_name: The name of the VM instance. device_name: The name of the disk device.
guest/disk/weighted_io_time ^GA IO Time
CUMULATIVE, INT64, ms gce_instance	The cumulative weighted IO time spent on the disk. Container-Optimized OS only. Sampled every 60 seconds. instance_name: The name of the VM instance. device_name: The name of the disk device.
guest/memory/anonymous_used ^BETA Anonymous memory usage in Bytes
GAUGE, INT64, By gce_instance	Anonymous memory usage, in Bytes. Summing values of all states yields the total anonymous memory used. Container-Optimized OS only. instance_name: The name of the VM instance. state: One of the following: active, inactive. Active means the memory has been used more recently.
guest/memory/bytes_used ^BETA Memory usage in Bytes
GAUGE, INT64, By gce_instance	Memory usage by each memory state, in Bytes. Summing values of all states yields the total memory on the machine. Container-Optimized OS only. instance_name: The name of the VM instance. state: One of the following: buffered, cached, free, slab, used.
guest/memory/dirty_used ^BETA Dirty pages usage in Bytes.
GAUGE, INT64, By gce_instance	Dirty pages usage, in Bytes. Container-Optimized OS only. instance_name: The name of the VM instance. state: One of the following: dirty, writeback. Dirty means the memory is waiting to be written back to disk, and writeback means the memory is actively being written back to disk.
guest/memory/page_cache_used ^BETA Page cache memory usage in Bytes
GAUGE, INT64, By gce_instance	Page cache memory usage, in Bytes. Summing values of all states yields the total anonymous memory used. Container-Optimized OS only. instance_name: The name of the VM instance. state: One of the following: active, inactive. Active means the memory has been used more recently and usually not reclaimed until needed.
guest/memory/unevictable_used ^BETA Unevictable memory usage in Bytes
GAUGE, INT64, By gce_instance	Unevictable memory usage, in Bytes. Container-Optimized OS only. instance_name: The name of the VM instance.
guest/system/os_feature_enabled ^BETA OS Feature
GAUGE, INT64, 1 gce_instance	OS Features like GPU support, KTD kernel, third party modules as unknown modules. 1 if the feature is enabled and 0, if disabled. Sampled every 60 seconds. instance_name: The name of the VM instance. value: Value for the features of the guest os system if present. os_feature: Features of the guest OS system.
guest/system/problem_count ^GA Problem Count
CUMULATIVE, INT64, 1 gce_instance	Number of times a machine problem has happened. Container-Optimized OS only. Sampled every 60 seconds. instance_name: The name of the VM instance. reason: Description of the type of the problem.
guest/system/problem_state ^GA Problem State
GAUGE, INT64, 1 gce_instance	Whether a problem is affecting the system or not. The problem is affecting the system when set to 1 and is not affecting the system when set to 0. Container-Optimized OS only. Sampled every 60 seconds. instance_name: The name of the VM instance. reason: Description of the cause of the problem. type: Description of the type of the problem.
guest/system/uptime ^GA Uptime
GAUGE, INT64, s gce_instance	Number of seconds that the operating system has been running for. Container-Optimized OS only. Sampled every 60 seconds. instance_name: The name of the VM instance. kernel_version: The version of the operating system kernel. os_version: The version of the operating system.
instance/cpu/reserved_cores ^GA Reserved vCPUs
GAUGE, DOUBLE, 1 gce_instance	Number of vCPUs reserved on the host of the instance. Sampled every 60 seconds. After sampling, data is not visible for up to 240 seconds. instance_name: The name of the VM instance.
instance/cpu/scheduler_wait_time ^GA Scheduler Wait Time
DELTA, DOUBLE, s{idle} gce_instance	Wait time is the time a vCPU is ready to run, but unexpectedly not scheduled to run. The wait time returned here is the accumulated value for all vCPUs. The time interval for which the value was measured is returned by Stackdriver in whole seconds as start_time and end_time. This metric is only available for VMs that belong to the e2 family. Sampled every 60 seconds. After sampling, data is not visible for up to 240 seconds. instance_name: The name of the VM instance.
instance/cpu/usage_time ^GA CPU usage
DELTA, DOUBLE, s{CPU} gce_instance	Delta vCPU usage for all vCPUs, in vCPU-seconds. To compute the per-vCPU utilization fraction, divide this value by (end-start)*N, where end and start define this value's time interval and N is `compute.googleapis.com/instance/cpu/reserved_cores` at the end of the interval. This value is reported by the hypervisor for the VM and can differ from `agent.googleapis.com/cpu/usage_time`, which is reported from inside the VM. Sampled every 60 seconds. After sampling, data is not visible for up to 240 seconds. instance_name: The name of the VM instance.
instance/cpu/utilization ^GA CPU utilization
GAUGE, DOUBLE, 10^2.% gce_instance	Fractional utilization of allocated CPU on this instance. Values are typically numbers between 0.0 and 1.0 (but some machine types allow bursting above 1.0). Charts display the values as a percentage between 0% and 100% (or more). This metric is reported by the hypervisor for the VM and can differ from `agent.googleapis.com/cpu/utilization`, which is reported from inside the VM. Sampled every 60 seconds. After sampling, data is not visible for up to 240 seconds. instance_name: The name of the VM instance.
instance/disk/max_read_bytes_count ^GA Peak disk read bytes
GAUGE, INT64, By gce_instance	Disk's maximum per-second read throughput over a period of time specified by the user. The period must be one minute or longer. Sampled every 60 seconds. After sampling, data is not visible for up to 120 seconds. device_name: The name of the disk device. storage_type: The storage type: `pd-standard` or `pd-ssd`. device_type: The disk type: `ephemeral` or `permanent`.
instance/disk/max_read_ops_count ^GA Peak disk read ops
GAUGE, INT64, 1 gce_instance	Disk's maximum per-second read requests count over a period of time specified by the user. The period must be one minute or longer. Sampled every 60 seconds. After sampling, data is not visible for up to 120 seconds. device_name: The name of the disk device. storage_type: The storage type: `pd-standard` or `pd-ssd`. device_type: The disk type: `ephemeral` or `permanent`.
instance/disk/max_write_bytes_count ^GA Peak disk write bytes
GAUGE, INT64, By gce_instance	Disk's maximum per-second write throughput over a period of time specified by the user. The period must be one minute or longer. Sampled every 60 seconds. After sampling, data is not visible for up to 120 seconds. device_name: The name of the disk device. storage_type: The storage type: `pd-standard` or `pd-ssd`. device_type: The disk type: `ephemeral` or `permanent`.
instance/disk/max_write_ops_count ^GA Peak disk write ops
GAUGE, INT64, 1 gce_instance	Disk's maximum per-second write requests count over a period of time specified by the user. The period must be one minute or longer. Sampled every 60 seconds. After sampling, data is not visible for up to 120 seconds. device_name: The name of the disk device. storage_type: The storage type: `pd-standard` or `pd-ssd`. device_type: The disk type: `ephemeral` or `permanent`.
instance/disk/read_bytes_count ^GA Disk read bytes
DELTA, INT64, By gce_instance	Delta count of bytes read from disk. Sampled every 60 seconds. After sampling, data is not visible for up to 240 seconds. instance_name: The name of the VM instance. device_name: The name of the disk device. storage_type: The storage type: `pd-standard`, `pd-ssd`, or `local-ssd`. device_type: The disk type: `ephemeral` or `permanent`.
instance/disk/read_ops_count ^GA Disk read operations
DELTA, INT64, 1 gce_instance	Delta count of disk read IO operations. Sampled every 60 seconds. After sampling, data is not visible for up to 240 seconds. instance_name: The name of the VM instance. device_name: The name of the disk device. storage_type: The storage type: `pd-standard`, `pd-ssd`, or `local-ssd`. device_type: The disk type: `ephemeral` or `permanent`.
instance/disk/throttled_read_bytes_count ^GA Throttled read bytes
DELTA, INT64, By gce_instance	Delta count of bytes in throttled read operations. Sampled every 60 seconds. After sampling, data is not visible for up to 240 seconds. instance_name: The name of the VM instance. device_name: The name of the disk device. storage_type: The storage type: `pd-standard` or `pd-ssd`. device_type: The disk type: `permanent`. throttle_reason: The limit that caused the throttling: `PER_GB`, `PER_VM`, or `EMERGENCY`. The `EMERGENCY` reason is deprecated but preserved for backward compatibility; the reported value is always 0.
instance/disk/throttled_read_ops_count ^GA Throttled read operations
DELTA, INT64, 1 gce_instance	Delta count of throttled read operations. Sampled every 60 seconds. After sampling, data is not visible for up to 240 seconds. instance_name: The name of the VM instance. device_name: The name of the disk device. storage_type: The storage type: `pd-standard` or `pd-ssd`. device_type: The disk type: `permanent`. throttle_reason: The limit that caused the throttling: `PER_GB`, `PER_VM`, or `EMERGENCY`. The `EMERGENCY` reason is deprecated but preserved for backward compatibility; the reported value is always 0.
instance/disk/throttled_write_bytes_count ^GA Throttled write bytes
DELTA, INT64, By gce_instance	Delta count of bytes in throttled write operations. Sampled every 60 seconds. After sampling, data is not visible for up to 240 seconds. instance_name: The name of the VM instance. device_name: The name of the disk device. storage_type: The storage type: `pd-standard` or `pd-ssd`. device_type: The disk type: `permanent`. throttle_reason: The limit that caused the throttling: `PER_GB`, `PER_VM`, or `EMERGENCY`. The `EMERGENCY` reason is deprecated but preserved for backward compatibility; the reported value is always 0.
instance/disk/throttled_write_ops_count ^GA Throttled write operations
DELTA, INT64, 1 gce_instance	Delta count of throttled write operations. Sampled every 60 seconds. After sampling, data is not visible for up to 240 seconds. instance_name: The name of the VM instance. device_name: The name of the disk device. storage_type: The storage type: `pd-standard` or `pd-ssd`. device_type: The disk type: `permanent`. throttle_reason: The limit that caused the throttling: `PER_GB`, `PER_VM`, or `EMERGENCY`. The `EMERGENCY` reason is deprecated but preserved for backward compatibility; the reported value is always 0.
instance/disk/write_bytes_count ^GA Disk write bytes
DELTA, INT64, By gce_instance	Delta count of bytes written to disk. Sampled every 60 seconds. After sampling, data is not visible for up to 240 seconds. instance_name: The name of the VM instance. device_name: The name of the disk device. storage_type: The storage type: `pd-standard`, `pd-ssd`, or `local-ssd`. device_type: The disk type: `ephemeral` or `permanent`.
instance/disk/write_ops_count ^GA Disk write operations
DELTA, INT64, 1 gce_instance	Delta count of disk write IO operations. Sampled every 60 seconds. After sampling, data is not visible for up to 240 seconds. instance_name: The name of the VM instance. device_name: The name of the disk device. storage_type: The storage type: `pd-standard`, `pd-ssd`, or `local-ssd`. device_type: The disk type: `ephemeral` or `permanent`.
instance/integrity/early_boot_validation_status ^BETA Early Boot Validation
GAUGE, INT64, 1 gce_instance	The validation status of early boot integrity policy. Sampled every 60 seconds. After sampling, data is not visible for up to 240 seconds. instance_name: The name of the VM instance. status: One of "passed", "failed", or "unknown".
instance/integrity/late_boot_validation_status ^BETA Late Boot Validation
GAUGE, INT64, 1 gce_instance	The validation status of late boot integrity policy. Sampled every 60 seconds. After sampling, data is not visible for up to 240 seconds. instance_name: The name of the VM instance. status: One of "passed", "failed", or "unknown".
instance/memory/balloon/ram_size ^GA VM Memory Total
GAUGE, INT64, By gce_instance	The total amount of memory in the VM. This metric is only available for VMs that belong to the e2 family. Sampled every 60 seconds. After sampling, data is not visible for up to 240 seconds. instance_name: The name of the VM instance.
instance/memory/balloon/ram_used ^GA VM Memory Used
GAUGE, INT64, By gce_instance	Memory currently used in the VM. This metric is only available for VMs that belong to the e2 family. Sampled every 60 seconds. After sampling, data is not visible for up to 240 seconds. instance_name: The name of the VM instance.
instance/memory/balloon/swap_in_bytes_count ^GA VM Swap In
DELTA, INT64, By gce_instance	The amount of memory read into the guest from its own swap space. This metric is only available for VMs that belong to the e2 family. Sampled every 60 seconds. After sampling, data is not visible for up to 240 seconds. instance_name: The name of the VM instance.
instance/memory/balloon/swap_out_bytes_count ^GA VM Swap Out
DELTA, INT64, By gce_instance	The amount of memory written from the guest to its own swap space. This metric is only available for VMs that belong to the e2 family. Sampled every 60 seconds. After sampling, data is not visible for up to 240 seconds. instance_name: The name of the VM instance.
instance/network/received_bytes_count ^GA Received bytes
DELTA, INT64, By gce_instance	Delta count of bytes received from the network. Sampled every 60 seconds. After sampling, data is not visible for up to 240 seconds. instance_name: The name of the VM instance. loadbalanced: (BOOL) Whether traffic was received by an L3 loadbalanced IP address assigned to the VM. Traffic that is externally routed to the VM's standard internal or external IP address, such as L7 loadbalanced traffic, is not considered to be loadbalanced in this metric.
instance/network/received_packets_count ^GA Received packets
DELTA, INT64, 1 gce_instance	Delta count of packets received from the network. Sampled every 60 seconds. After sampling, data is not visible for up to 240 seconds. instance_name: The name of the VM instance. loadbalanced: (BOOL) Whether traffic was received by an L3 loadbalanced IP address assigned to the VM. Traffic that is externally routed to the VM's standard internal or external IP address, such as L7 loadbalanced traffic, is not considered to be loadbalanced in this metric.
instance/network/sent_bytes_count ^GA Sent bytes
DELTA, INT64, By gce_instance	Delta count of bytes sent over the network. Sampled every 60 seconds. After sampling, data is not visible for up to 240 seconds. instance_name: The name of the VM instance. loadbalanced: (BOOL) Whether traffic was sent from an L3 loadbalanced IP address assigned to the VM. Traffic that is externally routed from the VM's standard internal or external IP address, such as L7 loadbalanced traffic, is not considered to be loadbalanced in this metric.
instance/network/sent_packets_count ^GA Sent packets
DELTA, INT64, 1 gce_instance	Delta count of packets sent over the network. Sampled every 60 seconds. After sampling, data is not visible for up to 240 seconds. instance_name: The name of the VM instance. loadbalanced: (BOOL) Whether traffic was sent from an L3 loadbalanced IP address assigned to the VM. Traffic that is externally routed from the VM's standard internal or external IP address, such as L7 loadbalanced traffic, is not considered to be loadbalanced in this metric.
instance/uptime ^GA Uptime
DELTA, DOUBLE, s{uptime} gce_instance	Delta of how long the VM has been running, in seconds. Note: to get the total number of seconds since VM start, use compute.googleapis.com/instance/uptime_total. Sampled every 60 seconds. After sampling, data is not visible for up to 240 seconds. instance_name: The name of the VM instance.
instance/uptime_total ^GA Uptime Total
GAUGE, INT64, s gce_instance	Elapsed time since the VM was started, in seconds. Sampled every 60 seconds. After sampling, data is not visible for up to 120 seconds. instance_name: The name of the VM instance.
instance_group/predicted_capacity ^BETA Instance group capacity as predicted
GAUGE, DOUBLE, 1 instance_group	Estimated capacity that the instance group will have if predictive autoscaling is enabled. Currently available only for autoscaled instance groups and equal to the predicted number of serving VMs multiplied by utilization target. Sampled every 60 seconds. After sampling, data is not visible for up to 240 seconds. metric_type: Metric for which utilization target is used, e.g., cpu.
instance_group/predicted_size ^BETA Instance group size as predicted
GAUGE, INT64, 1 instance_group	Estimated number of VMs that the instance group will have if predictive autoscaling is enabled. Currently available only for autoscaled instance groups. Sampled every 60 seconds. After sampling, data is not visible for up to 240 seconds. metric_type: Metric for which utilization target is used, e.g., cpu.
instance_group/predicted_utilization ^BETA Instance group utilization as predicted
GAUGE, DOUBLE, 1 instance_group	The sum of the estimated values of the scaling metric for all serving VMs if predictive autoscaling is enabled. Currently available only for autoscaled instance groups. Sampled every 60 seconds. After sampling, data is not visible for up to 240 seconds. metric_type: Metric for which utilization is predicted, e.g., cpu.
instance_group/size ^GA Instance group size
GAUGE, INT64, 1 instance_group	The number of VMs in the instance group. Sampled every 60 seconds. After sampling, data is not visible for up to 240 seconds.
mirroring/dropped_packets_count ^GA Dropped packets
DELTA, INT64, 1 gce_instance	Delta count of dropped mirrored packets. Sampled every 60 seconds. After sampling, data is not visible for up to 240 seconds. reason: The reason for the packet drop.
mirroring/mirrored_bytes_count ^GA Mirrored bytes
DELTA, INT64, By gce_instance	Delta count of mirrored bytes. Sampled every 60 seconds. After sampling, data is not visible for up to 240 seconds.
mirroring/mirrored_packets_count ^GA Mirrored packets
DELTA, INT64, 1 gce_instance	Delta count of mirrored packets. Sampled every 60 seconds. After sampling, data is not visible for up to 240 seconds.
nat/allocated_ports ^GA Allocated ports
GAUGE, INT64, {port} gce_instance	Number of ports allocated to a VM by the NAT gateway. Sampled every 60 seconds. After sampling, data is not visible for up to 165 seconds. nat_project_number: The project number to which the NAT gateway belongs. router_id: The Cloud Router ID to which the NAT gateway belongs. nat_gateway_name: The name of the NAT gateway. nat_ip: The NAT IP allocated to the NAT gateway.
nat/closed_connections_count ^GA Closed connections count
DELTA, INT64, {connection} gce_instance	Count of connections closed over the NAT gateway. Sampled every 60 seconds. After sampling, data is not visible for up to 165 seconds. nat_project_number: The project number to which the NAT gateway belongs. router_id: The Cloud Router ID to which the NAT gateway belongs. nat_gateway_name: The name of the NAT gateway. ip_protocol: The protocol of the connection. Can be TCP, UDP or ICMP.
nat/dropped_received_packets_count ^GA Received packets dropped count
DELTA, INT64, {packet} gce_instance	Count of received packets dropped by the NAT gateway. Sampled every 60 seconds. After sampling, data is not visible for up to 165 seconds. nat_project_number: The project number to which the NAT gateway belongs. router_id: The Cloud Router ID to which the NAT gateway belongs. nat_gateway_name: The name of the NAT gateway. ip_protocol: The protocol of the connection. Can be TCP, UDP or ICMP.
nat/dropped_sent_packets_count ^GA Sent packets dropped count
DELTA, INT64, {packet} gce_instance	Count of sent packets dropped by the NAT gateway. Sampled every 60 seconds. After sampling, data is not visible for up to 165 seconds. nat_project_number: The project number to which the NAT gateway belongs. router_id: The Cloud Router ID to which the NAT gateway belongs. nat_gateway_name: The name of the NAT gateway. ip_protocol: The protocol of the connection. Can be TCP, UDP or ICMP. reason: The reason for the packet drop. Possible values are OUT_OF_RESOURCES, ENDPOINT_INDEPENDENCE_CONFLICT.
nat/new_connections_count ^GA New connections count
DELTA, INT64, {connection} gce_instance	Count of new connections created over the NAT gateway. Sampled every 60 seconds. After sampling, data is not visible for up to 165 seconds. nat_project_number: The project number to which the NAT gateway belongs. router_id: The Cloud Router ID to which the NAT gateway belongs. nat_gateway_name: The name of the NAT gateway. ip_protocol: The protocol of the connection. Can be TCP, UDP or ICMP.
nat/open_connections ^GA Open connections
GAUGE, INT64, {connection} gce_instance	Number of connections currently open on the NAT gateway. Sampled every 60 seconds. After sampling, data is not visible for up to 165 seconds. nat_project_number: The project number to which the NAT gateway belongs. router_id: The Cloud Router ID to which the NAT gateway belongs. nat_gateway_name: The name of the NAT gateway. ip_protocol: The protocol of the connection. Can be TCP, UDP or ICMP.
nat/port_usage ^GA Port usage
GAUGE, INT64, {port} gce_instance	Maximum number of connections from a VM to a single internet endpoint (IP:port). Sampled every 60 seconds. After sampling, data is not visible for up to 165 seconds. nat_project_number: The project number to which the NAT gateway belongs. router_id: The Cloud Router ID to which the NAT gateway belongs. nat_gateway_name: The name of the NAT gateway. ip_protocol: The protocol of the connection. Can be TCP, UDP or ICMP.
nat/received_bytes_count ^GA Received bytes count
DELTA, INT64, By gce_instance	Count of bytes received (Internet -> VM) via the NAT gateway. Sampled every 60 seconds. After sampling, data is not visible for up to 165 seconds. nat_project_number: The project number to which the NAT gateway belongs. router_id: The Cloud Router ID to which the NAT gateway belongs. nat_gateway_name: The name of the NAT gateway. ip_protocol: The protocol of the connection. Can be TCP, UDP or ICMP.
nat/received_packets_count ^GA Received packets count
DELTA, INT64, {packet} gce_instance	Count of packets received (Internet -> VM) via the NAT gateway. Sampled every 60 seconds. After sampling, data is not visible for up to 165 seconds. nat_project_number: The project number to which the NAT gateway belongs. router_id: The Cloud Router ID to which the NAT gateway belongs. nat_gateway_name: The name of the NAT gateway. ip_protocol: The protocol of the connection. Can be TCP, UDP or ICMP.
nat/sent_bytes_count ^GA Sent bytes count
DELTA, INT64, By gce_instance	Count of bytes sent (VM -> Internet) over the NAT gateway. Sampled every 60 seconds. After sampling, data is not visible for up to 165 seconds. nat_project_number: The project number to which the NAT gateway belongs. router_id: The Cloud Router ID to which the NAT gateway belongs. nat_gateway_name: The name of the NAT gateway. ip_protocol: The protocol of the connection. Can be TCP, UDP or ICMP.
nat/sent_packets_count ^GA Sent packets count
DELTA, INT64, {packet} gce_instance	Count of packets sent (VM -> Internet) over the NAT gateway. Sampled every 60 seconds. After sampling, data is not visible for up to 165 seconds. nat_project_number: The project number to which the NAT gateway belongs. router_id: The Cloud Router ID to which the NAT gateway belongs. nat_gateway_name: The name of the NAT gateway. ip_protocol: The protocol of the connection. Can be TCP, UDP or ICMP.
quota/instances_per_vpc_network/exceeded ^ALPHA Instances Per VPC Network quota exceeded error
DELTA, INT64, 1 compute.googleapis.com/VpcNetwork	Number of attempts to exceed the limit on quota metric compute.googleapis.com/instances_per_vpc_network. limit_name: The limit name.
quota/instances_per_vpc_network/limit ^ALPHA Instances Per VPC Network quota limit
GAUGE, INT64, 1 compute.googleapis.com/VpcNetwork	Current limit on quota metric compute.googleapis.com/instances_per_vpc_network. Sampled every 60 seconds. limit_name: The limit name.
quota/instances_per_vpc_network/usage ^ALPHA Instances Per VPC Network quota usage
GAUGE, INT64, 1 compute.googleapis.com/VpcNetwork	Current usage on quota metric compute.googleapis.com/instances_per_vpc_network. limit_name: The limit name.
quota/internal_lb_forwarding_rules_per_vpc_network/exceeded ^ALPHA Internal Load Balancer Forwarding Rules per VPC Network quota exceeded error
DELTA, INT64, 1 compute.googleapis.com/VpcNetwork	Number of attempts to exceed the limit on quota metric compute.googleapis.com/internal_lb_forwarding_rules_per_vpc_network. limit_name: The limit name.
quota/internal_lb_forwarding_rules_per_vpc_network/limit ^ALPHA Internal Load Balancer Forwarding Rules per VPC Network quota limit
GAUGE, INT64, 1 compute.googleapis.com/VpcNetwork	Current limit on quota metric compute.googleapis.com/internal_lb_forwarding_rules_per_vpc_network. Sampled every 60 seconds. limit_name: The limit name.
quota/internal_lb_forwarding_rules_per_vpc_network/usage ^ALPHA Internal Load Balancer Forwarding Rules per VPC Network quota usage
GAUGE, INT64, 1 compute.googleapis.com/VpcNetwork	Current usage on quota metric compute.googleapis.com/internal_lb_forwarding_rules_per_vpc_network. limit_name: The limit name.
quota/internal_managed_forwarding_rules_per_vpc_network/exceeded ^ALPHA Internal Managed Forwarding Rules per VPC Network quota exceeded error
DELTA, INT64, 1 compute.googleapis.com/VpcNetwork	Number of attempts to exceed the limit on quota metric compute.googleapis.com/internal_managed_forwarding_rules_per_vpc_network. limit_name: The limit name.
quota/internal_managed_forwarding_rules_per_vpc_network/limit ^ALPHA Internal Managed Forwarding Rules per VPC Network quota limit
GAUGE, INT64, 1 compute.googleapis.com/VpcNetwork	Current limit on quota metric compute.googleapis.com/internal_managed_forwarding_rules_per_vpc_network. Sampled every 60 seconds. limit_name: The limit name.
quota/internal_managed_forwarding_rules_per_vpc_network/usage ^ALPHA Internal Managed Forwarding Rules per VPC Network quota usage
GAUGE, INT64, 1 compute.googleapis.com/VpcNetwork	Current usage on quota metric compute.googleapis.com/internal_managed_forwarding_rules_per_vpc_network. limit_name: The limit name.
quota/internal_protocol_forwarding_rules_per_vpc_network/exceeded ^ALPHA Internal Protocol Forwarding Rules per VPC Network quota exceeded error
DELTA, INT64, 1 compute.googleapis.com/VpcNetwork	Number of attempts to exceed the limit on quota metric compute.googleapis.com/internal_protocol_forwarding_rules_per_vpc_network. limit_name: The limit name.
quota/internal_protocol_forwarding_rules_per_vpc_network/limit ^ALPHA Internal Protocol Forwarding Rules per VPC Network quota limit
GAUGE, INT64, 1 compute.googleapis.com/VpcNetwork	Current limit on quota metric compute.googleapis.com/internal_protocol_forwarding_rules_per_vpc_network. Sampled every 60 seconds. limit_name: The limit name.
quota/internal_protocol_forwarding_rules_per_vpc_network/usage ^ALPHA Internal Protocol Forwarding Rules per VPC Network quota usage
GAUGE, INT64, 1 compute.googleapis.com/VpcNetwork	Current usage on quota metric compute.googleapis.com/internal_protocol_forwarding_rules_per_vpc_network. limit_name: The limit name.
quota/ip_aliases_per_vpc_network/exceeded ^ALPHA IP Aliases per VPC Network quota exceeded error
DELTA, INT64, 1 compute.googleapis.com/VpcNetwork	Number of attempts to exceed the limit on quota metric compute.googleapis.com/ip_aliases_per_vpc_network. limit_name: The limit name.
quota/ip_aliases_per_vpc_network/limit ^ALPHA IP Aliases per VPC Network quota limit
GAUGE, INT64, 1 compute.googleapis.com/VpcNetwork	Current limit on quota metric compute.googleapis.com/ip_aliases_per_vpc_network. Sampled every 60 seconds. limit_name: The limit name.
quota/ip_aliases_per_vpc_network/usage ^ALPHA IP Aliases per VPC Network quota usage
GAUGE, INT64, 1 compute.googleapis.com/VpcNetwork	Current usage on quota metric compute.googleapis.com/ip_aliases_per_vpc_network. limit_name: The limit name.
quota/psc_google_apis_forwarding_rules_per_vpc_network/exceeded ^ALPHA PSC Google APIs Forwarding Rules per VPC Network quota exceeded error
DELTA, INT64, 1 compute.googleapis.com/VpcNetwork	Number of attempts to exceed the limit on quota metric compute.googleapis.com/psc_google_apis_forwarding_rules_per_vpc_network. limit_name: The limit name.
quota/psc_google_apis_forwarding_rules_per_vpc_network/limit ^ALPHA PSC Google APIs Forwarding Rules per VPC Network quota limit
GAUGE, INT64, 1 compute.googleapis.com/VpcNetwork	Current limit on quota metric compute.googleapis.com/psc_google_apis_forwarding_rules_per_vpc_network. Sampled every 60 seconds. limit_name: The limit name.
quota/psc_google_apis_forwarding_rules_per_vpc_network/usage ^ALPHA PSC Google APIs Forwarding Rules per VPC Network quota usage
GAUGE, INT64, 1 compute.googleapis.com/VpcNetwork	Current usage on quota metric compute.googleapis.com/psc_google_apis_forwarding_rules_per_vpc_network. limit_name: The limit name.
quota/psc_ilb_consumer_forwarding_rules_per_producer_vpc_network/exceeded ^ALPHA PSC ILB Consumer Forwarding Rules per Producer VPC Network quota exceeded error
DELTA, INT64, 1 compute.googleapis.com/VpcNetwork	Number of attempts to exceed the limit on quota metric compute.googleapis.com/psc_ilb_consumer_forwarding_rules_per_producer_vpc_network. limit_name: The limit name.
quota/psc_ilb_consumer_forwarding_rules_per_producer_vpc_network/limit ^ALPHA PSC ILB Consumer Forwarding Rules per Producer VPC Network quota limit
GAUGE, INT64, 1 compute.googleapis.com/VpcNetwork	Current limit on quota metric compute.googleapis.com/psc_ilb_consumer_forwarding_rules_per_producer_vpc_network. Sampled every 60 seconds. limit_name: The limit name.
quota/psc_ilb_consumer_forwarding_rules_per_producer_vpc_network/usage ^ALPHA PSC ILB Consumer Forwarding Rules per Producer VPC Network quota usage
GAUGE, INT64, 1 compute.googleapis.com/VpcNetwork	Current usage on quota metric compute.googleapis.com/psc_ilb_consumer_forwarding_rules_per_producer_vpc_network. limit_name: The limit name.
quota/subnet_ranges_per_vpc_network/exceeded ^ALPHA Subnetwork ranges per VPC Network quota exceeded error
DELTA, INT64, 1 compute.googleapis.com/VpcNetwork	Number of attempts to exceed the limit on quota metric compute.googleapis.com/subnet_ranges_per_vpc_network. limit_name: The limit name.
quota/subnet_ranges_per_vpc_network/limit ^ALPHA Subnetwork ranges per VPC Network quota limit
GAUGE, INT64, 1 compute.googleapis.com/VpcNetwork	Current limit on quota metric compute.googleapis.com/subnet_ranges_per_vpc_network. Sampled every 60 seconds. limit_name: The limit name.
quota/subnet_ranges_per_vpc_network/usage ^ALPHA Subnetwork ranges per VPC Network quota usage
GAUGE, INT64, 1 compute.googleapis.com/VpcNetwork	Current usage on quota metric compute.googleapis.com/subnet_ranges_per_vpc_network. limit_name: The limit name.

container

Metrics from Google Kubernetes Engine. For newer Kubernetes metrics, see the kubernetes.io metrics. Launch stages of these metrics: GA

The "metric type" strings in this table must be prefixed with container.googleapis.com/. That prefix has been omitted from the entries in the table.

Metric type ^Launch stage Display name
Kind, Type, Unit Monitored resources	Description Labels
container/accelerator/duty_cycle ^GA Accelerator duty cycle
GAUGE, INT64, 1 gke_container	Percent of time over the past sample period (10s)during which the accelerator was actively processing. Sampled every 60 seconds. make: Make of the accelerator (e.g. nvidia) accelerator_id: ID of the accelerator. model: Model of the accelerator (e.g. 'Tesla P100')
container/accelerator/memory_total ^GA Accelerator memory total
GAUGE, INT64, By gke_container	Total accelerator memory in bytes. Sampled every 60 seconds. make: Make of the accelerator (e.g. nvidia) accelerator_id: ID of the accelerator. model: Model of the accelerator (e.g. 'Tesla P100')
container/accelerator/memory_used ^GA Accelerator memory used
GAUGE, INT64, By gke_container	Total accelerator memory allocated in bytes. Sampled every 60 seconds. make: Make of the accelerator (e.g. nvidia) accelerator_id: ID of the accelerator. model: Model of the accelerator (e.g. 'Tesla P100')
container/accelerator/request ^GA Request accelerators
GAUGE, INT64, {devices} gke_container	Number of accelerator devices requested by the container. Sampled every 60 seconds. resource_name: Name of the requested accelerator resource.
container/cpu/reserved_cores ^GA Reserved cores
GAUGE, DOUBLE, 1 gke_container	Number of cores of CPU reserved for the container. If no core limit is set, this will be zero. Sampled every 60 seconds.
container/cpu/usage_time ^GA CPU usage
CUMULATIVE, DOUBLE, s{CPU} gke_container	Cumulative CPU usage on all cores in seconds. This number divided by the elapsed time represents usage as a number of cores, regardless of any core limit that might be set. Sampled every 60 seconds.
container/cpu/utilization ^GA CPU utilization
GAUGE, DOUBLE, 10^2.% gke_container	The percentage of the allocated CPU that is currently in use on the container. Values are typically numbers between 0.0 and 1.0 (but might exceed 1.0), charts display the values as a percentage between 0% and 100% (or more). If no core limit is set, then this metric is not set. Sampled every 60 seconds. After sampling, data is not visible for up to 360 seconds.
container/disk/bytes_total ^GA Disk capacity
GAUGE, INT64, By gke_container	Total number of bytes of capacity on the disk identified by label ‘device_name’. Sampled every 60 seconds. device_name: The device name for the disk (e.g. `/dev/sda1`).
container/disk/bytes_used ^GA Disk usage
GAUGE, INT64, By gke_container	Total number of bytes used on the disk identified by label ‘device_name’. Sampled every 60 seconds. device_name: The device name for the disk (e.g. `/dev/sda1`).
container/disk/inodes_free ^GA Inode free
GAUGE, INT64, 1 gke_container	Free number of inodes on the disk identified by label 'device_name'. This metric is only for the `machine` container. Sampled every 60 seconds. device_name: The device name for the disk (e.g. `/dev/sda1`).
container/disk/inodes_total ^GA Inode capacity
GAUGE, INT64, 1 gke_container	Total number of inodes on the disk identified by label 'device_name'. This metric is only for the `machine` container. Sampled every 60 seconds. device_name: The device name for the disk (e.g. `/dev/sda1`).
container/memory/bytes_total ^GA Memory limit
GAUGE, INT64, By gke_container	Memory limit of the container in bytes. Sampled every 60 seconds.
container/memory/bytes_used ^GA Memory usage
GAUGE, INT64, By gke_container	Memory usage in bytes, broken down by type: evictable and non-evictable. Sampled every 60 seconds. memory_type: Either `evictable` or `non-evictable`. Evictable memory is memory that can be easily reclaimed by the kernel, while non-evictable memory cannot.
container/memory/page_fault_count ^GA Page faults
DELTA, INT64, 1 gke_container	Number of page faults, broken down by type: major and minor. Sampled every 60 seconds. After sampling, data is not visible for up to 360 seconds. fault_type: Fault type - either 'major' or 'minor', with the former indicating that the page had to be loaded from disk.
container/pid_limit ^GA PID capacity
GAUGE, INT64, 1 gke_container	The max PID of OS on the node. This metric is only for the `machine` container. Sampled every 60 seconds.
container/pid_used ^GA PID usage
GAUGE, INT64, 1 gke_container	The number of running process in the OS on the node. This metric is only for the `machine` container. Sampled every 60 seconds.
container/uptime ^GA Uptime
CUMULATIVE, DOUBLE, s gke_container	Number of seconds since the container started. Sampled every 60 seconds.

dataflow

Metrics from Dataflow. Launch stages of these metrics: ALPHA BETA EARLY_ACCESS GA

The "metric type" strings in this table must be prefixed with dataflow.googleapis.com/. That prefix has been omitted from the entries in the table.

Metric type ^Launch stage Display name
Kind, Type, Unit Monitored resources	Description Labels
job/backlog_bytes ^BETA Per-stage backlog in bytes
GAUGE, INT64, By dataflow_job	Amount of known, unprocessed input for a stage, in bytes. Sampled every 60 seconds. After sampling, data is not visible for up to 180 seconds. job_id: The ID of the current run of this pipeline. stage: The stage of the pipeline.
job/backlog_elements ^BETA Per-stage backlog in elements
GAUGE, INT64, 1 dataflow_job	Amount of known, unprocessed input for a stage, in elements. Sampled every 60 seconds. After sampling, data is not visible for up to 180 seconds. job_id: The ID of the current run of this pipeline. stage: The stage of the pipeline.
job/bigquery/write_count ^EARLY_ACCESS BigQueryIO.Write Requests
DELTA, INT64, 1 dataflow_job	BigQuery write requests from BigQueryIO.Write in Dataflow jobs. Sampled every 60 seconds. After sampling, data is not visible for up to 60 seconds. status: The canonical response code. See: https://cloud.google.com/apis/design/errors#handling_errors. job_id: The job id of the associated Dataflow job. ptransform: The pipeline PTransform making the requests. bigquery_project_id: The project of the associated BigQuery table, view or query. bigquery_dataset_id: The BigQuery dataset of the associated BigQuery table or view. bigquery_table_or_view_id: The BigQuery table or view id.
job/billable_shuffle_data_processed ^GA Billable shuffle data processed
GAUGE, INT64, By dataflow_job	The billable bytes of shuffle data processed by this Dataflow job. Sampled every 60 seconds. After sampling, data is not visible for up to 180 seconds. job_id: The ID of the current run of this pipeline.
job/current_num_vcpus ^GA Current number of vCPUs in use
GAUGE, INT64, 1 dataflow_job	The number of vCPUs currently being used by this Dataflow job. This is the current number of workers times the number of vCPUs per worker. Sampled every 60 seconds. After sampling, data is not visible for up to 180 seconds. job_id: The ID of the current run of this pipeline.
job/current_shuffle_slots ^BETA Current shuffle slots in use
GAUGE, INT64, 1 dataflow_job	The current shuffle slots used by this Dataflow job. Sampled every 60 seconds. After sampling, data is not visible for up to 180 seconds. job_id: The ID of the current run of this pipeline.
job/data_watermark_age ^GA Data watermark lag
GAUGE, INT64, s dataflow_job	The age (time since event timestamp) up to which all data has been processed by the pipeline. Sampled every 60 seconds. After sampling, data is not visible for up to 180 seconds. job_id: The ID of the current run of this pipeline.
job/elapsed_time ^GA Elapsed time
GAUGE, INT64, s dataflow_job	Duration that the current run of this pipeline has been in the Running state so far, in seconds. When a run completes, this stays at the duration of that run until the next run starts. Sampled every 60 seconds. After sampling, data is not visible for up to 180 seconds. job_id: The ID of the current run of this pipeline.
job/element_count ^GA Element count
GAUGE, INT64, 1 dataflow_job	Number of elements added to the pcollection so far. Sampled every 60 seconds. After sampling, data is not visible for up to 180 seconds. job_id: The ID of the current run of this pipeline. pcollection: The pipeline pcollection reporting this element count. PCollections are usually named by appending .out to the producing transform name.
job/elements_produced_count ^ALPHA Elements Produced
DELTA, INT64, 1 dataflow_job	The number of elements produced by each PTransform. Sampled every 60 seconds. After sampling, data is not visible for up to 180 seconds. job_id: The ID of the current run of this pipeline. pcollection: The pipeline PCollection to which the elements were produced. ptransform: The pipeline PTransform producing the elements.
job/estimated_byte_count ^GA Estimated byte count
GAUGE, INT64, By dataflow_job	An estimated number of bytes added to the pcollection so far. Dataflow calculates the average encoded size of elements in a pcollection and mutiplies it by the number of elements. Sampled every 60 seconds. After sampling, data is not visible for up to 180 seconds. job_id: The ID of the current run of this pipeline. pcollection: The pipeline pcollection reporting this estimated byte count. PCollections are usually named by appending .out to the producing transform name.
job/estimated_bytes_produced_count ^ALPHA Estimated Bytes Produced
DELTA, INT64, 1 dataflow_job	The estimated total byte size of elements produced by each PTransform. Sampled every 60 seconds. After sampling, data is not visible for up to 180 seconds. job_id: The ID of the current run of this pipeline. pcollection: The pipeline PCollection to which the elements were produced. ptransform: The pipeline PTransform producing the elements.
job/is_failed ^GA Failed
GAUGE, INT64, 1 dataflow_job	Has this job failed. Sampled every 60 seconds. After sampling, data is not visible for up to 60 seconds.
job/per_stage_data_watermark_age ^BETA Per-stage data watermark lag
GAUGE, INT64, s dataflow_job	The age (time since event timestamp) up to which all data has been processed by this stage of the pipeline. Sampled every 60 seconds. After sampling, data is not visible for up to 180 seconds. job_id: The ID of the current run of this pipeline. stage: The stage of the pipeline.
job/per_stage_system_lag ^BETA Per-stage system lag
GAUGE, INT64, s dataflow_job	The current maximum duration that an item of data has been processing or awaiting processing in seconds, per pipeline stage. Sampled every 60 seconds. After sampling, data is not visible for up to 180 seconds. job_id: The ID of the current run of this pipeline. stage: The stage of the pipeline.
job/pubsub/read_count ^BETA PubsubIO.Read requests from Dataflow jobs
DELTA, INT64, 1 dataflow_job	Pub/Sub Pull Requests. Sampled every 60 seconds. After sampling, data is not visible for up to 60 seconds. status: The canonical response code. See: https://cloud.google.com/apis/design/errors#handling_errors. job_id: The job id of the associated Dataflow job. ptransform: The pipeline PTransform making the requests. subscription_project_id: The project of the associated Pub/Sub subscription. subscription_id: The id of the associated Pub/Sub subscription. topic_project_id: The project of the associated Pub/Sub topic. topic_id: The id of the associated Pub/Sub topic.
job/pubsub/read_latencies ^BETA Pub/Sub Pull Request Latencies
DELTA, DISTRIBUTION, ms dataflow_job	Pub/Sub Pull request latencies from PubsubIO.Read in Dataflow jobs. Sampled every 60 seconds. After sampling, data is not visible for up to 60 seconds. job_id: The job id of the associated Dataflow job. ptransform: The pipeline PTransform making the requests. subscription_project_id: The project of the associated Pub/Sub subscription. subscription_id: The id of the associated Pub/Sub subscription. topic_project_id: The project of the associated Pub/Sub topic. topic_id: The id of the associated Pub/Sub topic.
job/pubsub/write_count ^BETA Pub/Sub Publish Requests
DELTA, INT64, 1 dataflow_job	Pub/Sub Publish requests from PubsubIO.Write in Dataflow jobs. Sampled every 60 seconds. After sampling, data is not visible for up to 60 seconds. status: The canonical response code. See: https://cloud.google.com/apis/design/errors#handling_errors. job_id: The job id of the associated Dataflow job. ptransform: The pipeline PTransform making the requests. topic_project_id: The project of the associated Pub/Sub topic. topic_id: The id of the associated Pub/Sub topic.
job/pubsub/write_latencies ^BETA Pub/Sub Publish Request Latencies
DELTA, DISTRIBUTION, ms dataflow_job	Pub/Sub Publish request latencies from PubsubIO.Write in Dataflow jobs. Sampled every 60 seconds. After sampling, data is not visible for up to 60 seconds. job_id: The job id of the associated Dataflow job. ptransform: The pipeline PTransform making the requests. topic_project_id: The project of the associated Pub/Sub topic. topic_id: The id of the associated Pub/Sub topic.
job/status ^GA Status
GAUGE, STRING,  dataflow_job	Current state of this pipeline (e.g.: RUNNING, DONE, CANCELLED, FAILED, ...). Not reported while the pipeline is not running. Sampled every 60 seconds. After sampling, data is not visible for up to 60 seconds. job_id: The ID of the current run of this pipeline.
job/system_lag ^GA System lag
GAUGE, INT64, s dataflow_job	The current maximum duration that an item of data has been processing or awaiting processing, in seconds. Sampled every 60 seconds. After sampling, data is not visible for up to 180 seconds. job_id: The ID of the current run of this pipeline.
job/total_memory_usage_time ^GA Total memory usage time
GAUGE, INT64, GBy.s dataflow_job	The total GB seconds of memory allocated to this Dataflow job. Sampled every 60 seconds. After sampling, data is not visible for up to 180 seconds. job_id: The ID of the current run of this pipeline.
job/total_pd_usage_time ^GA Total PD usage time
GAUGE, INT64, GBy.s dataflow_job	The total GB seconds for all persistent disk used by all workers associated with this Dataflow job. Sampled every 60 seconds. After sampling, data is not visible for up to 180 seconds. job_id: The ID of the current run of this pipeline. storage_type: The type of persistent disk: HDD, SSD.
job/total_shuffle_data_processed ^GA Total shuffle data processed
GAUGE, INT64, By dataflow_job	The total bytes of shuffle data processed by this Dataflow job. Sampled every 60 seconds. After sampling, data is not visible for up to 180 seconds. job_id: The ID of the current run of this pipeline.
job/total_streaming_data_processed ^GA Total streaming data processed
GAUGE, INT64, By dataflow_job	The total bytes of streaming data processed by this Dataflow job. Sampled every 60 seconds. After sampling, data is not visible for up to 180 seconds. job_id: The ID of the current run of this pipeline.
job/total_vcpu_time ^GA Total vCPU time
GAUGE, INT64, s dataflow_job	The total vCPU seconds used by this Dataflow job. Sampled every 60 seconds. After sampling, data is not visible for up to 180 seconds. job_id: The ID of the current run of this pipeline.
job/user_counter ^GA User Counter
GAUGE, DOUBLE, 1 dataflow_job	A user-defined counter metric. Sampled every 60 seconds. After sampling, data is not visible for up to 180 seconds. metric_name: The name provided by the user for this counter. job_id: The ID of the current run of this pipeline. ptransform: The pipeline PTransform reporting this counter value.

dataproc

Metrics from Dataproc. Launch stages of these metrics: BETA

The "metric type" strings in this table must be prefixed with dataproc.googleapis.com/. That prefix has been omitted from the entries in the table.

Metric type ^Launch stage Display name
Kind, Type, Unit Monitored resources	Description Labels
cluster/hdfs/datanodes ^BETA HDFS DataNodes
GAUGE, INT64, 1 cloud_dataproc_cluster	Indicates the number of HDFS DataNodes that are running inside a cluster. Sampled every 60 seconds. After sampling, data is not visible for up to 120 seconds. status: The status of DataNode such as running, decommissioning, decommissioned.
cluster/hdfs/storage_capacity ^BETA HDFS capacity
GAUGE, DOUBLE, GiBy cloud_dataproc_cluster	Indicates capacity of HDFS system running on cluster in GB. Sampled every 60 seconds. After sampling, data is not visible for up to 120 seconds. status: The status of HDFS capacity such as used, remaining.
cluster/hdfs/storage_utilization ^BETA HDFS storage utilization
GAUGE, DOUBLE, 1 cloud_dataproc_cluster	The percentage of HDFS storage currently used. Sampled every 60 seconds. After sampling, data is not visible for up to 120 seconds.
cluster/hdfs/unhealthy_blocks ^BETA Unhealthy HDFS blocks by status
GAUGE, INT64, 1 cloud_dataproc_cluster	Indicates the number of unhealthy blocks inside the cluster. Sampled every 60 seconds. After sampling, data is not visible for up to 120 seconds. status: The status of unhealthy HDFS block such as missing, under_replication, corrupt, pending_deletion, missing_repl_one.
cluster/job/completion_time ^BETA Job duration
DELTA, DISTRIBUTION, s cloud_dataproc_cluster	The time jobs took to complete from the time the user submits a job to the time Dataproc reports it is completed. Sampled every 60 seconds. After sampling, data is not visible for up to 120 seconds. job_type: The type of job such as HADOOP_JOB or SPARK_JOB.
cluster/job/duration ^BETA Job state duration
DELTA, DISTRIBUTION, s cloud_dataproc_cluster	The time jobs have spent in a given state. Sampled every 60 seconds. After sampling, data is not visible for up to 120 seconds. job_type: The type of job such as HADOOP_JOB or SPARK_JOB. state: The state of the job such as PENDING or RUNNING.
cluster/job/failed_count ^BETA Failed jobs
DELTA, INT64, 1 cloud_dataproc_cluster	Indicates the number of jobs that have failed on a cluster. Sampled every 60 seconds. After sampling, data is not visible for up to 120 seconds. error_type: The rpc code of the error such as 404 or 503. job_type: The type of job such as HADOOP_JOB or SPARK_JOB.
cluster/job/running_count ^BETA Running jobs
GAUGE, INT64, 1 cloud_dataproc_cluster	Indicates the number of jobs that are running on a cluster. Sampled every 60 seconds. After sampling, data is not visible for up to 120 seconds. job_type: The type of job such as HADOOP_JOB or SPARK_JOB.
cluster/job/submitted_count ^BETA Submitted jobs
DELTA, INT64, 1 cloud_dataproc_cluster	Indicates the number of jobs that have been submitted to a cluster. Sampled every 60 seconds. After sampling, data is not visible for up to 120 seconds. job_type: The type of job such as HADOOP_JOB or SPARK_JOB.
cluster/operation/completion_time ^BETA Operation duration
DELTA, DISTRIBUTION, s cloud_dataproc_cluster	The time operations took to complete from the time the user submits a operation to the time Dataproc reports it is completed. Sampled every 60 seconds. After sampling, data is not visible for up to 120 seconds. operation_type: The type of operation such as CREATE_CLUSTER or DELETE_CLUSTER.
cluster/operation/duration ^BETA Operation state duration
DELTA, DISTRIBUTION, s cloud_dataproc_cluster	The time operations have spent in a given state. Sampled every 60 seconds. After sampling, data is not visible for up to 120 seconds. operation_type: The state of the operation such as PENDING or CREATE_VMS. state: The type of operation such as CREATE_CLUSTER or DELETE_CLUSTER.
cluster/operation/failed_count ^BETA Failed operations
DELTA, INT64, 1 cloud_dataproc_cluster	Indicates the number of operations that have failed on a cluster. Sampled every 60 seconds. After sampling, data is not visible for up to 120 seconds. error_type: The rpc code of the error such as 404 or 503. operation_type: The type of operation such as CREATE_CLUSTER or DELETE_CLUSTER.
cluster/operation/running_count ^BETA Running operations
GAUGE, INT64, 1 cloud_dataproc_cluster	Indicates the number of operations that are running on a cluster. Sampled every 60 seconds. After sampling, data is not visible for up to 120 seconds. operation_type: The type of operation such as CREATE_CLUSTER or DELETE_CLUSTER.
cluster/operation/submitted_count ^BETA Submitted operations
DELTA, INT64, 1 cloud_dataproc_cluster	Indicates the number of operations that have been submitted to a cluster. Sampled every 60 seconds. After sampling, data is not visible for up to 120 seconds. operation_type: The type of operation such as CREATE_CLUSTER or DELETE_CLUSTER.
cluster/yarn/allocated_memory_percentage ^BETA YARN allocated memory percentage
GAUGE, DOUBLE, 1 cloud_dataproc_cluster	The percentage of YARN memory is allocated. Sampled every 60 seconds. After sampling, data is not visible for up to 120 seconds.
cluster/yarn/apps ^BETA YARN active applications
GAUGE, INT64, 1 cloud_dataproc_cluster	Indicates the number of active YARN applications. Sampled every 60 seconds. After sampling, data is not visible for up to 120 seconds. status: The status of YARN application such as running, pending, completed, failed, killed.
cluster/yarn/containers ^BETA YARN containers
GAUGE, INT64, 1 cloud_dataproc_cluster	Indicates the number of YARN containers. Sampled every 60 seconds. After sampling, data is not visible for up to 120 seconds. status: The status of YARN container such as allocated, pending, reserved.
cluster/yarn/memory_size ^BETA YARN memory size
GAUGE, DOUBLE