Ops Agent metrics

When you install the Ops Agent on your VM instances, the agent transmits data for the metric types listed on this page to Monitoring. For information about choosing the right agent, see Ops Agent overview.

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

  • 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.

  • The metric types in the table titled "agent" are free; the other types are chargeable. For pricing information, see Monitoring pricing.

Metrics from the Ops Agent and the Monitoring agent

The Monitoring agent has two entirely different implementations, one for Linux VMs and one for Windows VMs. The Ops Agent runs on both Linux and Windows VMs.

Linux only?

Some of the metrics listed in the tables on this page include the disclaimer "Linux only". This disclaimer means that the Monitoring agent for Windows VMs does not collect this metric. The disclaimer does not apply to the Ops Agent running on Windows VMs.

Differences in metric collection

The following table describes differences in the data ingested by the Ops Agent and the Monitoring agent.
Metric type, omitting
agent.googleapis.com
Ops Agent (GA) Ops Agent (Preview) Monitoring agent
disk/bytes_used and
disk/percent_used
Ingested with the full path in the device label; for example, /dev/sd15.

Not ingested for virtual devices like tmpfs and udev.
Ingested without /dev in the path in the device label; for example, sda15.

Ingested for virtual devices like tmpfs and udev.
Ingested without /dev in the path in the device label; for example, sda15.

Ingested for virtual devices like tmpfs and udev.
processes/count_by_state Not ingested. Ingested. Ingested.
swap/bytes_used Ingested without the device label on Linux. Ingested with the device label on Linux. Ingested with the device label on Linux.
The GA column refers to Ops Agent versions 2.0.0 and higher. The Preview column refers to Ops Agent versions less than 2.0.0.

Agent metrics

Metrics from the Ops Agent running on VM instances in Google Cloud.

agent

Metrics from the default configuration for the Ops Agent. Launch stages of these metrics: GA

The "metric type" strings in this table must be prefixed with agent.googleapis.com/agent/. 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
Monitoring Agent API Request Count
CUMULATIVEINT641
aws_ec2_instance, gae_app, gce_instance, k8s_container
Count of API requests from the Monitoring agent. Linux only. Sampled every 60 seconds.
state: Request state.
memory_usage GA
Monitoring Agent Memory Usage
GAUGEDOUBLEBy
aws_ec2_instance, gae_app, gce_instance, k8s_container
Memory in use by the Monitoring agent. Linux only. Sampled every 60 seconds.
monitoring/point_count GA
Monitoring Agent Metric Point Count
CUMULATIVEINT641
aws_ec2_instance, gae_app, gce_instance, k8s_container
Count of metric points written to Stackdriver Monitoring. Linux only. Sampled every 60 seconds.
status: gRPC response code.
uptime GA
Monitoring/Logging Agent Uptime
CUMULATIVEINT64s
aws_ec2_instance, gae_app, gce_instance, k8s_container
Uptime of Monitoring agent or Logging agent. Linux only. Sampled every 60 seconds.
version: Agent version.

Table built at 2021-10-06 17:04:40 UTC.

cpu

Metrics from the default configuration for the Ops Agent and the Monitoring agent. Compute Engine also reports similar metrics, but those come from the hypervisor rather than from inside the VM. Launch stages of these metrics: GA

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

Metric type Launch stage
Display name
Kind, Type, Unit
Monitored resources
Description
Labels
load_15m GA
CPU load (15m)
GAUGEDOUBLE1
aws_ec2_instance, gce_instance
Load average over the last 15 minutes. Linux only. Sampled every 60 seconds.
load_1m GA
CPU load (1m)
GAUGEDOUBLE1
aws_ec2_instance, gce_instance
Load average over the last 1 minute. Linux only. Sampled every 60 seconds.
load_5m GA
CPU load (5m)
GAUGEDOUBLE1
aws_ec2_instance, gce_instance
Load average over the last 5 minutes. Linux only. Sampled every 60 seconds.
usage_time GA
CPU usage time (OS reported)
CUMULATIVEINT64s{CPU}
aws_ec2_instance, gce_instance
Total time the CPU spent in different states. This value is reported from inside the VM and can differ from `compute.googleapis.com/instance/cpu/usage_time`, which is reported by the hypervisor for the VM. Linux only. Sampled every 60 seconds.
cpu_number: CPU number, for example, "0", "1", or "2". This label is only set with certain Monitoring configurations.
cpu_state: CPU state, one of [idle, interrupt, nice, softirq, steal, system, user, wait].
utilization GA
CPU utilization (OS reported)
GAUGEDOUBLE%
aws_ec2_instance, gce_instance
Percentage of the total CPU capacity spent in different states. This value is reported from inside the VM and can differ from `compute.googleapis.com/instance/cpu/utilization`, which is reported by the hypervisor for the VM. Sampled every 60 seconds.
cpu_number: CPU number, for example, "0", "1", or "2". This label is only set with certain Monitoring configurations. Linux only.
cpu_state: CPU state, one of [idle, interrupt, nice, softirq, steal, system, user, wait] on Linux or [idle, used] on Windows.

disk

Metrics from the default configuration for the Ops Agent and the Monitoring agent. Launch stages of these metrics: GA

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

Metric type Launch stage
Display name
Kind, Type, Unit
Monitored resources
Description
Labels
bytes_used GA
Disk bytes used
GAUGEDOUBLEBy
aws_ec2_instance, gce_instance
Current number of disk bytes used by state. Summing the values of all states yields the total available disk space. Linux only. Sampled every 60 seconds.
device: Device name.
state: Type of usage, one of [free, used, reserved].
io_time GA
Disk I/O time
CUMULATIVEINT64ms
aws_ec2_instance, gce_instance
Cumulative time spent while I/O operations are in queue and when the disks are busy. The value of this metric increases when there is at least one I/O operation pending or in progress. This value corresponds to Field 10 in the https://www.kernel.org/doc/Documentation/iostats.txt documentation. Linux only. Sampled every 60 seconds.
device: Device name.
merged_operations GA
Disk merged operations
CUMULATIVEINT641
aws_ec2_instance, gce_instance
Total count of merged disk operations by direction. Linux only. Sampled every 60 seconds.
device: Device name.
direction: Direction, one of [read, write].
operation_count GA
Disk operations
CUMULATIVEINT641
aws_ec2_instance, gce_instance
Total count of disk operations by direction. Linux only. Sampled every 60 seconds.
device: Device name.
direction: Direction, one of [read, write].
operation_time GA
Disk operation time
CUMULATIVEINT64ms
aws_ec2_instance, gce_instance
Average time to complete a disk operation multiplied by wall clock time. When aligned as a rate, the value is the average time to complete a disk operation. Linux only. Sampled every 60 seconds.
device: Device name.
direction: Direction, one of [read, write].
pending_operations GA
Disk pending operations
GAUGEDOUBLE1
aws_ec2_instance, gce_instance
Current count of pending disk operations. Linux only. Sampled every 60 seconds.
device: Device name.
percent_used GA
Disk utilization
GAUGEDOUBLE%
aws_ec2_instance, gce_instance
Current percentage of the disk used by state. The utilization is between 0.0 and 100.0. Sampled every 60 seconds.
device: Device name.
state: Type of usage, one of [free, used, reserved].
read_bytes_count GA
Disk bytes read
CUMULATIVEINT64By
aws_ec2_instance, gce_instance
Total number of disk bytes that have been read. Linux only. Sampled every 60 seconds.
device: Device name.
weighted_io_time GA
Disk weighted I/O time
CUMULATIVEINT64ms
aws_ec2_instance, gce_instance
Total time spent on disk I/O operations weighted by the number of in-progress I/Os. This metric provides a measure of the I/O completion time and of the size of the backlog that might be accumulating. Linux only. Sampled every 60 seconds.
device: Device name.
write_bytes_count GA
Disk bytes written
CUMULATIVEINT64By
aws_ec2_instance, gce_instance
Total number of disk bytes that have been written. Linux only. Sampled every 60 seconds.
device: Device name.

iis

Microsoft Windows only. Metrics from the default configuration for the Microsoft Windows Ops Agent and the Microsoft Windows Monitoring agent. Launch stages of these metrics: GA

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

Metric type Launch stage
Display name
Kind, Type, Unit
Monitored resources
Description
Labels
current_connections GA
IIS open connections
GAUGEDOUBLE1
aws_ec2_instance, gce_instance
Currently open connections to IIS. Sampled every 60 seconds.
network/transferred_bytes_count GA
IIS transferred bytes
CUMULATIVEINT64By
aws_ec2_instance, gce_instance
Network bytes transferred by IIS. Sampled every 60 seconds.
direction: Direction (sent, received)
new_connection_count GA
IIS connections
CUMULATIVEINT641
aws_ec2_instance, gce_instance
Connections opened to IIS. Sampled every 60 seconds.
request_count GA
IIS requests
CUMULATIVEINT641
aws_ec2_instance, gce_instance
Requests made to IIS. Sampled every 60 seconds.
http_method: Http method (post, put, get, delete, option, trace, head)

interface

Network-interface metrics from the default configuration for the Ops Agent and the Linux Monitoring agent. Launch stages of these metrics: GA

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

Metric type Launch stage
Display name
Kind, Type, Unit
Monitored resources
Description
Labels
errors GA
Network errors
CUMULATIVEINT641
aws_ec2_instance, gce_instance
Total count of network errors. Linux only. Sampled every 60 seconds.
device: Name of the network device.
direction: Direction of the network traffic, one of [tx, rx].
packets GA
Network packets
CUMULATIVEINT641
aws_ec2_instance, gce_instance
Total count of packets sent over the network. Linux only. Sampled every 60 seconds.
device: Name of the network device.
direction: Direction of the network traffic, one of [tx, rx].
traffic GA
Network traffic
CUMULATIVEINT64By
aws_ec2_instance, gce_instance
Total number of bytes sent over the network. Linux only. Sampled every 60 seconds.
device: Name of the network device.
direction: Direction of the network traffic, one of [tx, rx].

memory

Metrics from the default configuration for the Ops Agent and the Monitoring agent. Launch stages of these metrics: GA

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

Metric type Launch stage
Display name
Kind, Type, Unit
Monitored resources
Description
Labels
bytes_used GA
Memory usage
GAUGEDOUBLEBy
aws_ec2_instance, gce_instance
Current memory usage by memory state. Summing the values of all states yields the total memory on the machine. Linux only. Sampled every 60 seconds.
state: One of [buffered, cached, free, slab, used].
percent_used GA
Memory utilization
GAUGEDOUBLE%
aws_ec2_instance, gce_instance
Current percentage of memory used by memory state. Summing percentages over all states yields 100 percent. Sampled every 60 seconds.
state: One of [buffered, cached, free, slab, used].

mssql

Microsoft Windows only. Metrics from the default configuration for the Microsoft Windows Ops Agent and the Microsoft Windows Monitoring agent. Launch stages of these metrics: GA

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

Metric type Launch stage
Display name
Kind, Type, Unit
Monitored resources
Description
Labels
connections/user GA
SQL Server open connections
GAUGEDOUBLE1
aws_ec2_instance, gce_instance
Currently open connections to SQL server. Sampled every 60 seconds.
transaction_rate GA
SQL Server transaction rate
GAUGEDOUBLE1/s
aws_ec2_instance, gce_instance
SQL server total transactions per second. Sampled every 60 seconds.
write_transaction_rate GA
SQL Server write transaction rate
GAUGEDOUBLE1/s
aws_ec2_instance, gce_instance
SQL server write transactions per second. Sampled every 60 seconds.

network

Metrics from the default configuration for the Ops Agent and the Monitoring agent. Launch stages of these metrics: GA

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

Metric type Launch stage
Display name
Kind, Type, Unit
Monitored resources
Description
Labels
tcp_connections GA
TCP connections
GAUGEDOUBLE1
aws_ec2_instance, gce_instance
Current count of TCP connections. Sampled every 60 seconds.
port: TCP port number on the local or remote host, for example, "0-local", "0-remote", or "all".
tcp_state: TCP state, for example, "ESTABLISHED" or "FIN_WAIT1".

pagefile

Metrics from the default configuration for the Ops Agent and the Microsoft Windows Monitoring agent. Launch stages of these metrics: GA

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

Metric type Launch stage
Display name
Kind, Type, Unit
Monitored resources
Description
Labels
percent_used GA
Pagefile utilization
GAUGEDOUBLE%
aws_ec2_instance, gce_instance
Current percentage of pagefile used by state. Values are between 0.0 and 100.0. Windows only. Sampled every 60 seconds.
state: Type of usage, one of [free, used].

processes

Metrics from the default configuration for the Ops Agent and the Monitoring agent. Launch stages of these metrics: GA

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

Metric type Launch stage
Display name
Kind, Type, Unit
Monitored resources
Description
Labels
count_by_state GA
Processes
GAUGEDOUBLE1
aws_ec2_instance, gce_instance
Count of processes in the given state. Linux only. Sampled every 60 seconds.
state: Running, sleeping, zombie, etc.
cpu_time GA
Process CPU
CUMULATIVEINT64us{CPU}
aws_ec2_instance, gce_instance
CPU time of the given process. Sampled every 60 seconds.
process: Process name.
user_or_syst: Whether a user or system process.
command: Process command.
command_line: Process command line, 1024 characters maximum.
owner: Process owner.
pid: Process ID.
disk/read_bytes_count GA
Process disk read I/O
CUMULATIVEINT64By
aws_ec2_instance, gce_instance
Process disk read I/O. Linux only. Sampled every 60 seconds.
process: Process name.
command: Process command.
command_line: Process command line, 1024 characters maximum.
owner: Process owner.
pid: Process ID.
disk/write_bytes_count GA
Process disk write I/O
CUMULATIVEINT64By
aws_ec2_instance, gce_instance
Process disk write I/O. Linux only. Sampled every 60 seconds.
process: Process name.
command: Process command.
command_line: Process command line, 1024 characters maximum.
owner: Process owner.
pid: Process ID.
fork_count GA
Fork count
CUMULATIVEINT641
aws_ec2_instance, gce_instance
Total number of processes forked. Linux only. Sampled every 60 seconds.
rss_usage GA
Process resident memory
GAUGEDOUBLEBy
aws_ec2_instance, gce_instance
Resident memory usage of the given process. Linux only. Sampled every 60 seconds.
process: Process name.
command: Process command.
command_line: Process command line, 1024 characters maximum.
owner: Process owner.
pid: Process ID.
vm_usage GA
Process virtual memory
GAUGEDOUBLEBy
aws_ec2_instance, gce_instance
VM Usage of the given process. Sampled every 60 seconds.
process: Process name.
command: Process command.
command_line: Process command line, 1024 characters maximum.
owner: Process owner.
pid: Process ID.

swap

Metrics from the default configuration for the Ops Agent and the Linux Monitoring agent. Launch stages of these metrics: GA

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

Metric type Launch stage
Display name
Kind, Type, Unit
Monitored resources
Description
Labels
bytes_used GA
Swap usage
GAUGEDOUBLEBy
aws_ec2_instance, gce_instance
Current number of swap bytes used by state. For this metric to be available, swap must be enabled on the instance. Linux only. Sampled every 60 seconds.
device: Swap device.
state: One of [used, free, cached, reserved].
io GA
Swap I/O operations
CUMULATIVEINT641
aws_ec2_instance, gce_instance
Total count of swap I/O operations by direction. For this metric to be available, swap must be enabled on the instance. Linux only. Sampled every 60 seconds.
direction: Swap I/O direction, one of [in, out].
percent_used GA
Swap utilization
GAUGEDOUBLE%
aws_ec2_instance, gce_instance
Current percentage of swap used by state. Values are between 0.0 and 100.0. For this metric to be available, swap must be enabled on the instance. Linux only. Sampled every 60 seconds.
device: Swap device.
state: One of [used, free, cached, reserved].

Third-party application metrics

JVM

Metrics for JVM. For information about collecting these metrics, see the guide for configuring the Ops Agent for JVM.

Metric type 
Kind, Type
Monitored resources
Labels
workload.googleapis.com/jvm.classes.loaded
GAUGEINT64
gce_instance
 
workload.googleapis.com/jvm.gc.collections.count
CUMULATIVEINT64
gce_instance
name
workload.googleapis.com/jvm.gc.collections.elapsed
CUMULATIVEINT64
gce_instance
name
workload.googleapis.com/jvm.memory.heap.committed
GAUGEINT64
gce_instance
 
workload.googleapis.com/jvm.memory.heap.init
GAUGEINT64
gce_instance
 
workload.googleapis.com/jvm.memory.heap.max
GAUGEINT64
gce_instance
 
workload.googleapis.com/jvm.memory.heap.used
GAUGEINT64
gce_instance
 
workload.googleapis.com/jvm.memory.nonheap.committed
GAUGEINT64
gce_instance
 
workload.googleapis.com/jvm.memory.nonheap.init
GAUGEINT64
gce_instance
 
workload.googleapis.com/jvm.memory.nonheap.max
GAUGEINT64
gce_instance
 
workload.googleapis.com/jvm.memory.nonheap.used
GAUGEINT64
gce_instance
 
workload.googleapis.com/jvm.memory.pool.committed
GAUGEINT64
gce_instance
name
workload.googleapis.com/jvm.memory.pool.init
GAUGEINT64
gce_instance
name
workload.googleapis.com/jvm.memory.pool.max
GAUGEINT64
gce_instance
name
workload.googleapis.com/jvm.memory.pool.used
GAUGEINT64
gce_instance
name
workload.googleapis.com/jvm.threads.count
GAUGEINT64
gce_instance
 

nginx

Metrics for nginx. For information about collecting these metrics, see the guide for configuring the Ops Agent for nginx.

Metric type 
Kind, Type
Monitored resources
Labels
workload.googleapis.com/nginx.requests
CUMULATIVEINT64
gce_instance
 
workload.googleapis.com/nginx.connections_accepted
CUMULATIVEINT64
gce_instance
 
workload.googleapis.com/nginx.connections_handled
CUMULATIVEINT64
gce_instance
 
workload.googleapis.com/nginx.connections_current
GAUGEINT64
gce_instance
state