OpenStack (legacy)

Supported OS Linux Windows Mac OS

Versión de la integración4.0.0
This page is not yet available in Spanish. We are working on its translation.
If you have any questions or feedback about our current translation project, feel free to reach out to us!

OpenStack default dashboard

Overview

Note: This integration only applies to OpenStack v12 and below. If you are looking to collect metrics from OpenStack v13+, use the OpenStack Controller integration.

Get metrics from OpenStack service in real time to:

  • Visualize and monitor OpenStack states.
  • Be notified about OpenStack failovers and events.

Setup

Installation

To capture your OpenStack metrics, install the Agent on your hosts running hypervisors.

Configuration

Prepare OpenStack

Configure a Datadog role and user with your identity server:

openstack role create datadog_monitoring
openstack user create datadog \
    --password my_password \
    --project my_project_name
openstack role add datadog_monitoring \
    --project my_project_name \
    --user datadog

Then, update your policy.json files to grant the needed permissions. role:datadog_monitoring requires access to the following operations:

Nova

{
  "compute_extension": "aggregates",
  "compute_extension": "hypervisors",
  "compute_extension": "server_diagnostics",
  "compute_extension": "v3:os-hypervisors",
  "compute_extension": "v3:os-server-diagnostics",
  "compute_extension": "availability_zone:detail",
  "compute_extension": "v3:availability_zone:detail",
  "compute_extension": "used_limits_for_admin",
  "os_compute_api:os-aggregates:index": "rule:admin_api or role:datadog_monitoring",
  "os_compute_api:os-aggregates:show": "rule:admin_api or role:datadog_monitoring",
  "os_compute_api:os-hypervisors": "rule:admin_api or role:datadog_monitoring",
  "os_compute_api:os-server-diagnostics": "rule:admin_api or role:datadog_monitoring",
  "os_compute_api:os-used-limits": "rule:admin_api or role:datadog_monitoring"
}

Neutron

{
  "get_network": "rule:admin_or_owner or rule:shared or rule:external or rule:context_is_advsvc or role:datadog_monitoring"
}

Keystone

{
  "identity:get_project": "rule:admin_required or project_id:%(target.project.id)s or role:datadog_monitoring",
  "identity:list_projects": "rule:admin_required or role:datadog_monitoring"
}

You may need to restart your Keystone, Neutron, and Nova API services to ensure that the policy changes take.

Note: Installing the OpenStack integration could increase the number of VMs that Datadog monitors. For more information on how this may affect your billing, see the Billing FAQ.

Agent configuration

  1. Configure the Datadog Agent to connect to your Keystone server, and specify individual projects to monitor. Edit the openstack.d/conf.yaml file in the conf.d/ folder at the root of your Agent’s configuration directory with the configuration below. See the sample openstack.d/conf.yaml for all available configuration options:

    init_config:
      ## @param keystone_server_url - string - required
      ## Where your identity server lives.
      ## Note that the server must support Identity API v3
      #
      keystone_server_url: "https://<KEYSTONE_SERVER_ENDPOINT>:<PORT>/"
    
    instances:
      ## @param name - string - required
      ## Unique identifier for this instance.
      #
      - name: "<INSTANCE_NAME>"
    
        ## @param user - object - required
        ## User credentials
        ## Password authentication is the only auth method supported.
        ## `user` object expects the parameter `username`, `password`,
        ## and `user.domain.id`.
        ##
        ## `user` should resolve to a structure like:
        ##
        ##  {'password': '<PASSWORD>', 'name': '<USERNAME>', 'domain': {'id': '<DOMAINE_ID>'}}
        #
        user:
          password: "<PASSWORD>"
          name: datadog
          domain:
            id: "<DOMAINE_ID>"
    
  2. Restart the Agent.

Log collection
  1. Collecting logs is disabled by default in the Datadog Agent, you can enable it in datadog.yaml:

    logs_enabled: true
    
  2. Add this configuration block to your openstack.d/conf.yaml file to start collecting your Openstack logs:

    logs:
      - type: file
        path: "<LOG_FILE_PATH>"
        source: openstack
    

    Change the path parameter value and configure them for your environment. See the sample openstack.d/conf.yaml for all available configuration options.

Validation

Run the Agent’s status subcommand and look for openstack under the Checks section.

Data Collected

Metrics

openstack.nova.current_workload
(gauge)
Current workload on the Nova hypervisor
openstack.nova.disk_available_least
(gauge)
Disk available for the Nova hypervisor
Shown as gibibyte
openstack.nova.free_disk_gb
(gauge)
Free disk on the Nova hypervisor
Shown as gibibyte
openstack.nova.free_ram_mb
(gauge)
Free RAM on the Nova hypervisor
Shown as mebibyte
openstack.nova.hypervisor_load.1
(gauge)
The average hypervisor load over one minute.
openstack.nova.hypervisor_load.15
(gauge)
The average hypervisor load over fifteen minutes.
openstack.nova.hypervisor_load.5
(gauge)
The average hypervisor load over five minutes.
openstack.nova.limits.max_image_meta
(gauge)
The maximum allowed image metadata definitions for this tenant
openstack.nova.limits.max_personality
(gauge)
The maximum allowed personalities for this tenant
openstack.nova.limits.max_personality_size
(gauge)
The maximum size of a single personality allowed for this tenant
openstack.nova.limits.max_security_group_rules
(gauge)
The maximum number of security group rules allowed for this tenant
openstack.nova.limits.max_security_groups
(gauge)
The maximum number of security groups allowed for this tenant
openstack.nova.limits.max_server_meta
(gauge)
The maximum allowed service metadata definitions for this tenant
openstack.nova.limits.max_total_cores
(gauge)
The maximum allowed cores for this tenant
openstack.nova.limits.max_total_floating_ips
(gauge)
The maximum allowed floating IPs for this tenant
openstack.nova.limits.max_total_instances
(gauge)
The maximum number of instances allowed for this tenant
openstack.nova.limits.max_total_keypairs
(gauge)
The maximum allowed key pairs allowed for this tenant
openstack.nova.limits.max_total_ram_size
(gauge)
The max allowed RAM size for this tenant in megabytes (MB)
Shown as mebibyte
openstack.nova.limits.total_cores_used
(gauge)
The total cores used by this tenant
openstack.nova.limits.total_floating_ips_used
(gauge)
The total floating IPs used by this tenant
openstack.nova.limits.total_instances_used
(gauge)
The total instances used by this tenant
openstack.nova.limits.total_ram_used
(gauge)
The current RAM used by this tenant in megabytes (MB)
Shown as mebibyte
openstack.nova.limits.total_security_groups_used
(gauge)
The total number of security groups used by this tenant
openstack.nova.local_gb
(gauge)
The size in GB of the ephemeral disk present on this hypervisor host
Shown as gibibyte
openstack.nova.local_gb_used
(gauge)
The size in GB of disk used on this hypervisor host
Shown as gibibyte
openstack.nova.memory_mb
(gauge)
The size in MB of RAM present on this hypervisor host
Shown as mebibyte
openstack.nova.memory_mb_used
(gauge)
The size in MB of RAM used on this hypervisor host
Shown as mebibyte
openstack.nova.running_vms
(gauge)
Number of running VMs on this hypervisor host
openstack.nova.server.cpu0_time
(gauge)
CPU time in nanoseconds of this virtual CPU
Shown as nanosecond
openstack.nova.server.hdd_errors
(gauge)
The number of errors seen by the server when accessing an HDD device
openstack.nova.server.hdd_read
(gauge)
Number of bytes read from an HDD device by this server
Shown as byte
openstack.nova.server.hdd_read_req
(gauge)
The number of read requests made to an HDD device by this server
openstack.nova.server.hdd_write
(gauge)
Number of bytes written to an HDD device by this server
Shown as byte
openstack.nova.server.hdd_write_req
(gauge)
The number of write requests made to an HDD device by this server
openstack.nova.server.memory
(gauge)
The amount of memory in MB provisioned for this server
Shown as mebibyte
openstack.nova.server.memory_actual
(gauge)
The amount of memory in MB provisioned for this server
Shown as mebibyte
openstack.nova.server.memory_rss
(gauge)
The amount of memory used by the processes of this server that is not associated with disk pages - such as stack and heap memory
Shown as mebibyte
openstack.nova.server.vda_errors
(gauge)
The number of errors seen by the server when accessing a VDA device
openstack.nova.server.vda_read
(gauge)
Number of bytes read from a VDA device by this server
Shown as byte
openstack.nova.server.vda_read_req
(gauge)
The number of read requests made to a VDA device by this server
openstack.nova.server.vda_write
(gauge)
Number of bytes written to a VDA device by this server
Shown as byte
openstack.nova.server.vda_write_req
(gauge)
The number of write requests made to a VDA device by this server
openstack.nova.vcpus
(gauge)
Number of vCPUs available on this hypervisor host
openstack.nova.vcpus_used
(gauge)
Number of vCPUS used on this hypervisor host

Events

The OpenStack check does not include any events.

Service Checks

openstack.neutron.api.up
Returns CRITICAL if the Agent is unable to query the Neutron API, UNKNOWN if there is an issue with the Keystone API. Returns OK otherwise.
Statuses: ok, critical, unknown

openstack.nova.api.up
Returns CRITICAL if the Agent is unable to query the Nova API, UNKNOWN if there is an issue with the Keystone API. Returns OK otherwise.
Statuses: ok, critical, unknown

openstack.keystone.api.up
Returns CRITICAL if the Agent is unable to query the Keystone API. Returns OK otherwise.
Statuses: ok, critical

openstack.nova.hypervisor.up
Returns UNKNOWN if the Agent is unable to get the Hypervisor state, CRITICAL if the Hypervisor is down. Returns OK otherwise.
Statuses: ok, critical, unknown

openstack.neutron.network.up
Returns UNKNOWN if the Agent is unable to get the Network state, CRITICAL if the Network is down. Returns OK otherwise.
Statuses: ok, critical, unknown

Troubleshooting

Need help? Contact Datadog support.

Further Reading

Additional helpful documentation, links, and articles:

PREVIEWING: rtrieu/product-analytics-ui-changes