The Ops Agent is the primary agent for collecting telemetry from yourCompute Engine instances. Combining the collection of logs, metrics, andtraces into a single process, the Ops Agent usesFluent Bitfor logs, which supports high-throughput logging, and theOpenTelemetryCollector formetrics and traces.

For information about installing the Ops Agent, see the following:

• For Compute Engine instances, seeInstalling the Ops Agent.
• For Bare Metal Solution servers, seeSet up Ops Agent for Bare Metal Solution.

Ops Agent features

Overall features include:

• Single download and installation/upgrade process.
• Simple, unified, YAML-based configuration.
• Support forstandard Linux and Windows distros.
• Proxy support.

Logging features

Logging features include:

• Improved performance compared to the legacy Logging Agent:

  • High throughput capability, taking full advantage of multi-core architecture.
  • Efficient resource (e.g. memory, CPU) management.

• Collecting logs fromvarious sources:

  • Standard system logs (/var/log/syslog and/var/log/messages for Linux,Windows Event Log) collected with no configuration.
  • File-based logs with customizable paths and refresh interval.
  • Journald daemon / systemd logs.
  • Logs over TCP protocol.
  • Logs over Forward protocol (used by Fluent Bit and Fluentd).

• Flexible processing:

  • Parse text logs into structured logs:JSON-based andregular-expression-based parsing.
  • Modifylog entries by removing, renaming, or setting fields.
  • Excludelogs based on labels and regular expressions.
  • Detect and concatenatemultilinelanguage-exception logs from Java, Python, and Golang.

• Third-party application support

  • Curated third-party application log integration that recognizes common app logfile paths and formats.

Monitoring features

Monitoring features include:

• System metrics collected with no configuration. Metrics collected include:
  • cpu metrics
  • disk metrics
  • iis metrics (Windows only)
  • interface metrics
  • gpu metrics (Linux only)
  • memory metrics
  • mssql metrics (Windows only)
  • pagefile metrics (Windowsonly)
  • swap metrics
  • network metrics
  • processes metrics
  • agent self metrics:
• Third-party application support
  • Curatedintegrations for third-party applicationmetrics, whichcollect common app metrics and offer sample dashboards and alert policies.

• Collection of Prometheus metrics from applications running on Compute Engine.
• Collection of OpenTelemetry Protocol (OTLP) metrics and traces from applications instrumented withOpenTelemetry SDKs.
• Collection of NVIDIA Data Center GPU Manager (DCGM) metrics.

Note: Some metrics are only available on VMs running on Linux. See detailedmetrics definition and whether they are Linux only in theagentmetrics page.

The Cloud MonitoringMetrics Management page provides informationthat can help you control the amount you spend on billable metricswithout affecting observability. TheMetrics Management page reports thefollowing information:

• Ingestion volumes for both byte- and sample-based billing, across metric domains and for individual metrics.
• Data about labels and cardinality of metrics.
• Number of reads for each metric.
• Use of metrics in alerting policies and custom dashboards.
• Rate of metric-write errors.

You can also use theMetrics Management page toexclude unneeded metrics,eliminating the cost of ingesting them.For more information about theMetrics Management page, seeView and manage metric usage.

Trace features

Trace features include:

• Collection ofOpenTelemetry Protocol (OTLP) traces fromapplications instrumented withOpenTelemetry SDKs.

Virtual machine instances

You can install the Ops Agent onCompute Engine instances.To create a Compute Engine instance, see theCompute Engine getting started guide.

Note:The Ops Agent is not supported on Amazon Elastic Compute Cloud (Amazon EC2)instances or on-premises machines.

The following types of VM instances belong to managed services that implementservice-specificMonitoringsupport. Don't try to manually install or configure the Ops Agent on them:

• App Engine standard has built-in Monitoring support. Agents aren't needed.
• App Engine flexible environment instances have pre-installed agents with service-specific configurations.
• Dataflow instances have pre-installed legacy Monitoring agents with service-specific configurations.

• Dataproc instances prior to image version 2.2 have pre-installed legacy Monitoring agents with service-specific configurations.

  You can install the Ops Agent on Dataproc clusters that use image version 2.2 and later to collect syslog logs and host metrics. For more information, see Dataproc 2.2.x release versions.

• Google Kubernetes Engine node instances:
  • GKE on Google Cloud clusters are configured tocollect logs and metrics with Logging and Monitoring by default. You can also configure metric and log collection for existing container clusters. For information, seeConfiguring logging and monitoring for GKE.
  • Google Distributed Cloud instances have an integrated logging and monitoring solution that collects status about system components. For information, see Logging and monitoring.

Operating systems

The Ops Agent supports the following operating systems on compatible VMinstances. For information about Arm VM instances, seeSupport for Compute Engine Arm VMs.

Linux operating systems

The agent supports the following Linux operating systems:

Note: To use Cloud Logging on VMs with Container-Optimized OS, followUsing Cloud Logging with Container-Optimized OS instead.

• Rocky Linux 8
• Rocky Linux 9
• RHEL 8: rhel-8, rhel-8-6-sap-ha, rhel-8-8-sap-ha, rhel-8-10-sap-ha
• RHEL 9: rhel-9
• Debian 11 (Bullseye)
• Debian 12 (Bookworm)
• Deep Learning VM Images based on Debian 11 (Bullseye)
• Ubuntu 22.04 LTS (Jammy Jellyfish): ubuntu-2204-lts, ubuntu-minimal-2204-lts
• Ubuntu 24.04 LTS (Noble Numbat): ubuntu-2404-lts-amd64, ubuntu-minimal-2404-lts-amd64
• SLES 12: sles-12, sles-12-sp5-sap
• SLES 15: sles-15, sles-15-sp3-sap, sles-15-sp4-sap, sles-15-sp5-sap,sles-15-sp6-sap
• OpenSUSE Leap 15: opensuse-leap (opensuse-leap-15-6-*)

Support for Compute Engine Arm VMs

The Ops Agent supports Compute EngineC4A and Tau T2AArm VM instances. The Ops Agent issupported for Arm VMs on the following OS images:

• Debian 12 (Bookworm): Ops Agent version 2.44.0
• Ubuntu 22.04 LTS (Jammy Jellyfish): Ops Agent version 2.42.0
• Ubuntu 24.04 LTS (Noble Numbat): Ops Agent version 2.47.0
• RHEL 9: Ops Agent version 2.40.0
• Rocky Linux 8: Ops Agent version 2.51.0
• Rocky Linux 9: Ops Agent version 2.40.0
• SLES 15: Ops Agent version 2.43.0
• OpenSUSE Leap 15: Ops Agent version 2.43.0

The Ops Agent is supported on a subset of the images supported byCompute Engine. For more information about Compute Engine Arm VMs,seeCreating and starting an Arm VMinstance.

Windows operating systems

Note:n1-standard-1 virtual machines are too small to run Windows and theOps Agent, so you must use a larger VM.

The agent supports the following Windows operating systems:

• Windows Server: windows-2016, windows-2019, windows-2022, windows-2025
• Windows Server Core: windows-2016-core, windows-2019-core,windows-20h2-core, windows-2022-core, windows-2025-core

Agent access requirements

Installing the agent requires access to the following DNS names:

• Google Cloud package repository:packages.cloud.google.com

• Google downloads subdomain:dl.google.com

Running the agent requires access to the following DNS names:

• OAuth2 token server:oauth2.googleapis.com

  Earlier versions of the agent may require access towww.googleapis.com(full URL:https://www.googleapis.com/oauth2/v3/token).

  If you're using an older version of the agent, it's recommended that youupgrade your agent to the latestversion.

• Monitoring APIs:monitoring.googleapis.com

• Logging APIs:logging.googleapis.com

The agent uses TCP ports within the VM for inter-process communication. Thefollowing ports, which don't need to be open externally, must be available tothe agent on your VM:

• 20201

• 20202

Support policy

The Ops Agent is supported according to thesupport time horizons policy.

Pricing

If you install the Ops Agent, then you might be charged for the metrics,logs, or traces that the agent sends to your Google Cloud project. For pricinginformation, seeGoogle Cloud Observability pricing.

To collect OpenTelemetry Protocol (OTLP) traces, you must configure theagent to collect traces from an instrumented application; for moreinformation, seeCollect (OTLP) traces with the Ops Agent.

What's next

• To install the Ops Agent, seeInstalling theOps Agent.