Export Metrics From a CockroachDB Dedicated Cluster

On this page Carat arrow pointing down

CockroachDB Dedicated users can use the Cloud API to configure metrics export to AWS CloudWatch, Datadog, or Prometheus. Once the export is configured, metrics will flow from all nodes in all regions of your CockroachDB Dedicated cluster to your chosen cloud metrics sink.

Tip:

To export metrics from a CockroachDB Self-Hosted cluster, refer to Monitoring and Alerting instead of this page.

Exporting metrics to AWS CloudWatch is only available on CockroachDB Dedicated clusters which are hosted on AWS. Metrics export to Datadog and Prometheus is supported on all CockroachDB Dedicated clusters.

Note:

This feature is in preview. This feature is subject to change. To share feedback and/or issues, contact Support.

metricexport endpoint

To configure and manage metrics export for your CockroachDB Dedicated cluster, use the metricexport endpoint appropriate for your desired cloud metrics sink:

Cloud metrics sink Metrics export endpoint
AWS Cloudwatch https://cockroachlabs.cloud/api/v1/clusters/{your_cluster_id}/metricexport/cloudwatch
Datadog https://cockroachlabs.cloud/api/v1/clusters/{your_cluster_id}/metricexport/datadog
Prometheus https://cockroachlabs.cloud/api/v1/clusters/{your_cluster_id}/metricexport/prometheus

Access to the metricexport endpoints requires a valid CockroachDB Cloud service account with the appropriate permissions (admin privilege, Cluster Administrator role, or Cluster Operator role).

The following methods are available for use with the metricexport endpoints, and require the listed service account permissions:

Method Required permissions Description
GET ADMIN, EDIT, or READ Returns the current status of the metrics export configuration.
POST ADMIN or EDIT Enables metrics export, or updates an existing metrics export configuration.
DELETE ADMIN Disables metrics export, halting all metrics export to AWS CloudWatch, Datadog, or Prometheus.

See Service accounts for instructions on configuring a service account with these required permissions.

Enable metrics export

Exporting metrics to AWS CloudWatch is only available on CockroachDB Dedicated clusters which are hosted on AWS. If your CockroachDB Dedicated cluster is hosted on GCP or Azure, you can export metrics to Datadog or Prometheus instead.

Note:

Enabling metrics export will send around 250 metrics per node to AWS CloudWatch. Review the AWS CloudWatch documentation to gauge how this adds to your AWS CloudWatch spend.

Perform the following steps to enable metrics export from your CockroachDB Dedicated cluster to AWS CloudWatch.

  1. Create the desired target AWS CloudWatch log group by following the Create a log group in CloudWatch instructions. If you already have a log group created, you may skip this step. When your CockroachDB Dedicated cluster emits metrics to AWS CloudWatch, they are written to this log group.

  2. Find your CockroachDB Dedicated cluster ID:

    1. Visit the CockroachDB Cloud console cluster page.
    2. Click on the name of your cluster.
    3. Find your cluster ID in the URL of the single cluster overview page: https://cockroachlabs.cloud/cluster/{your_cluster_id}/overview. The ID should resemble f78b7feb-b6cf-4396-9d7f-494982d7d81e.
  3. Determine your cluster's cloud provider account ID. This command uses the third-party JSON parsing tool jq to isolate just the needed account_id field:

    icon/buttons/copy
    curl --request GET \
      --url https://cockroachlabs.cloud/api/v1/clusters/{your_cluster_id} \
      --header 'Authorization: Bearer {secret_key}' | jq .account_id
    

    Where:

    • {your_cluster_id} is the cluster ID of your CockroachDB Dedicated cluster as determined in step 2.
    • {secret_key} is your CockroachDB Dedicated API key. See API Access for more details.
  4. Create a cross-account IAM role in your AWS account:

    1. In the AWS console, visit the IAM page.
    2. Select Roles and click Create role.
    3. For Trusted entity type, select AWS account.
    4. Choose Another AWS account.
    5. For Account ID, provide the CockroachDB Dedicated cloud provider account ID from step 3.
    6. Finish creating the IAM role with a suitable name. These instructions will use the role name CockroachCloudMetricsExportRole. You do not need to add any permissions.
    Note:

    You will need the Amazon Resource Name (ARN) for your cross-account IAM role later in this procedure.

  5. Select the new role, and create a new policy for this role. These instructions will use the policy name CockroachCloudMetricsExportPolicy.

  6. Select the new policy, and paste the following into the Permissions tab, with the {} JSON option selected:

    icon/buttons/copy
    {
        "Version": "2012-10-17",
        "Statement": [
            {
                "Action": [
                    "logs:CreateLogGroup",
                    "logs:CreateLogStream",
                    "logs:DescribeLogGroups",
                    "logs:DescribeLogStreams",
                    "logs:PutRetentionPolicy",
                    "logs:PutLogEvents"
                ],
                "Effect": "Allow",
                "Resource": [
                    "arn:aws:logs:*:{your_aws_acct_id}:log-group:{log_group_name}",
                    "arn:aws:logs:*:{your_aws_acct_id}:log-group:{log_group_name}:log-stream:*"
                ]
            }
        ]
    }
    

    Where:

    • {your_aws_acct_id} is the AWS Account ID of the AWS account where you created the CockroachCloudMetricsExportRole role, not the cloud provider account ID of your CockroachDB Dedicated cluster from step 3. You can find your AWS Account ID on the AWS IAM page.
    • {log_group_name} is the target AWS CloudWatch log group you created in step 1.

    This defines the set of permissions that the CockroachDB Dedicated metrics export feature requires to be able to write metrics to CloudWatch.

  7. Copy the Amazon Resource Name (ARN) of the CockroachCloudMetricsExportRole role found under Summary, which is needed for the next step.

  8. Issue the following Cloud API command to enable metrics export for your CockroachDB Dedicated cluster:

    icon/buttons/copy
    curl --request POST \
      --url https://cockroachlabs.cloud/api/v1/clusters/{cluster_id}/metricexport/cloudwatch \
      --header "Authorization: Bearer {secret_key}" \
      --data '{"target_region": "{aws_region}", "role_arn": "arn:aws:iam::{role_arn}:role/CockroachCloudMetricsExportRole", "log_group_name": "{log_group_name}"}'
    

    Where:

    • {cluster_id} is your CockroachDB Dedicated cluster ID as determined in step 2.
    • {secret_key} is your CockroachDB Dedicated API key. See API Access for instructions on generating this key.
    • {aws_region} is your AWS region, like us-east-1.
    • {role_arn} is the ARN for the CockroachCloudMetricsExportRole role you copied in step 7. If you used a different role name there, be sure to use your role name in place of CockroachCloudMetricsExportRole in the above command.
    • {log_group_name} is the target AWS CloudWatch log group you created in step 1. This must be the same group name you provided in step 6.

    Specifying an AWS region (to {aws_region}) that you do not have a cluster in, or a region that only partially covers your cluster's nodes will result in missing metrics.

  9. Depending on the size of your cluster and how many regions it spans, the configuration may take a moment. You can monitor the ongoing status of the configuration using the following Cloud API command:

    icon/buttons/copy
    curl --request GET \
      --url https://cockroachlabs.cloud/api/v1/clusters/{cluster_id}/metricexport/cloudwatch \
      --header "Authorization: Bearer {secret_key}"
    

    Run the command periodically until the command returns a status of ENABLED, at which point the configuration across all nodes is complete, and metrics will begin appearing in CloudWatch under the log group you created in step 1. Since the configuration is applied to cluster nodes in a rolling fashion, you may see some metrics appear even before the GET command returns an ENABLED status.

  10. Once metrics export has been enabled, you can access metrics from your CockroachDB Dedicated cluster directly in AWS CloudWatch.

To enable metrics export for your CockroachDB Dedicated cluster to Datadog, you can either enable the Datadog integration in your CockroachDB Dedicated Cloud Console, or on the command line via the Cloud API:

  1. Find your CockroachDB Dedicated cluster ID:

    1. Visit the CockroachDB Cloud console cluster page.
    2. Click on the name of your cluster.
    3. Find your cluster ID in the URL of the single cluster overview page: https://cockroachlabs.cloud/cluster/{your_cluster_id}/overview. The ID should resemble f78b7feb-b6cf-4396-9d7f-494982d7d81e.
  2. Determine the Datadog API key you'd like to use. If you don't already have one, follow the steps to add a new Datadog API key.

  3. Issue the following Cloud API command to enable metrics export for your CockroachDB Dedicated cluster:

    icon/buttons/copy
    curl --request POST \
      --url https://cockroachlabs.cloud/api/v1/clusters/{cluster_id}/metricexport/datadog \
      --header "Authorization: Bearer {secret_key}" \
      --data '{"site": "{datadog_site}", "api_key": "{datadog_api_key}"}'
    

    Where:

    • {cluster_id} is your CockroachDB Dedicated cluster ID as determined in step 1, resembling f78b7feb-b6cf-4396-9d7f-494982d7d81e.
    • {secret_key} is your CockroachDB Dedicated API key. See API Access for instructions on generating this key.
    • {datadog_site} is your Datadog site. Valid sites are: US1, US3, US5, US1_GOV, and EU1.
    • {datadog_api_key} is the Datadog API key determined in step 2.
  4. Depending on the size of your cluster and how many regions it spans, the configuration may take a moment. You can monitor the ongoing status of the configuration using the following Cloud API command:

    icon/buttons/copy
    curl --request GET \
      --url https://cockroachlabs.cloud/api/v1/clusters/{cluster_id}/metricexport/datadog \
      --header "Authorization: Bearer {secret_key}"
    

    Run the command periodically until the command returns a status of ENABLED, at which point the configuration across all nodes is complete, and metrics will begin appearing in the Datadog interface. Since the configuration is applied to cluster nodes in a rolling fashion, you may see some metrics appear even before the GET command returns an ENABLED status.

  5. Once metrics export has been enabled, you can access metrics from your CockroachDB Dedicated cluster directly in Datadog's Metrics Explorer, or via Datadog's notebook or dashboard features.

  6. Review enable percentiles for selected metrics. Configure metrics as necessary.

Note:

A subset of CockroachDB metrics require that you explicitly enable percentiles for them in the Datadog interface. Graphs that display data for these metrics will fail to render properly otherwise.

  1. Find your CockroachDB Dedicated cluster ID:

    1. Visit the CockroachDB Cloud console cluster page.
    2. Click on the name of your cluster.
    3. Find your cluster ID in the URL of the single cluster overview page: https://cockroachlabs.cloud/cluster/{your_cluster_id}/overview. The ID should resemble f78b7feb-b6cf-4396-9d7f-494982d7d81e.
  2. Issue the following Cloud API command to enable metrics export for your CockroachDB Dedicated cluster:

    icon/buttons/copy
    curl --request POST \
      --url https://cockroachlabs.cloud/api/v1/clusters/{cluster_id}/metricexport/prometheus \
      --header "Authorization: Bearer {secret_key}"
    

    Where:

    • {cluster_id} is your CockroachDB Dedicated cluster ID as determined in step 1, resembling f78b7feb-b6cf-4396-9d7f-494982d7d81e.
    • {secret_key} is your CockroachDB Dedicated API key. See API Access for instructions on generating this key.
  3. Depending on the size of your cluster and how many regions it spans, the configuration may take a moment. You can monitor the ongoing status of the configuration using the following Cloud API command:

    icon/buttons/copy
    curl --request GET \
      --url https://cockroachlabs.cloud/api/v1/clusters/{cluster_id}/metricexport/prometheus \
      --header "Authorization: Bearer {secret_key}"
    

    Run the command periodically until the command returns a status of ENABLED, at which point the configuration across all nodes is complete. The response also includes targets, a map of scrape endpoints exposing metrics to regions. For example:

    {
      "cluster_id": "f78b7feb-b6cf-4396-9d7f-494982d7d81e",
      "user_message": "This integration is active.",
      "status": "ENABLED",
      "targets": {
        "us-east4": "https://cockroachlabs.cloud/api/v1/clusters/f78b7feb-b6cf-4396-9d7f-494982d7d81e/metricexport/prometheus/us-east4/scrape"
      }
    }
    

    There is a separate scrape endpoint per region if you have a multi-region cluster.

  4. You can test a scrape endpoint by using the following Cloud API command:

    icon/buttons/copy
    curl --request GET \
      --url https://cockroachlabs.cloud/api/v1/clusters/{cluster_id}/metricexport/prometheus/{cluster_region}/scrape \
      --header "Authorization: Bearer {secret_key}"
    

    Where:

    • {cluster_id} is your CockroachDB Dedicated cluster ID as determined in step 1, resembling f78b7feb-b6cf-4396-9d7f-494982d7d81e.
    • {cluster_region} is a region of your CockroachDB Dedicated cluster as shown in the targets of step 3, such as us-east4. You can also find your cluster’s region(s) on the Cluster Overview page.
    • {secret_key} is your CockroachDB Dedicated API key. See API Access for instructions on generating this key.

    Metrics are labeled with the cluster name and ID, node, organization name, and region. The beginning lines of a metrics scrape response follows:

    # HELP crdb_dedicated_addsstable_applications Number of SSTable ingestions applied (i.e. applied by Replicas)
    # TYPE crdb_dedicated_addsstable_applications counter
    crdb_dedicated_addsstable_applications{cluster="test-gcp",instance="172.28.0.167:8080",node="cockroachdb-j5t6j",node_id="1",organization="CRL - Test",region="us-east4",store="1"} 0
    crdb_dedicated_addsstable_applications{cluster="test-gcp",instance="172.28.0.49:8080",node="cockroachdb-ttzj8",node_id="3",organization="CRL - Test",region="us-east4",store="3"} 0
    crdb_dedicated_addsstable_applications{cluster="test-gcp",instance="172.28.0.99:8080",node="cockroachdb-r5rns",node_id="2",organization="CRL - Test",region="us-east4",store="2"} 0
    # HELP crdb_dedicated_addsstable_copies number of SSTable ingestions that required copying files during application
    # TYPE crdb_dedicated_addsstable_copies counter
    crdb_dedicated_addsstable_copies{cluster="test-gcp",instance="172.28.0.167:8080",node="cockroachdb-j5t6j",node_id="1",organization="CRL - Test",region="us-east4",store="1"} 0
    crdb_dedicated_addsstable_copies{cluster="test-gcp",instance="172.28.0.49:8080",node="cockroachdb-ttzj8",node_id="3",organization="CRL - Test",region="us-east4",store="3"} 0
    crdb_dedicated_addsstable_copies{cluster="test-gcp",instance="172.28.0.99:8080",node="cockroachdb-r5rns",node_id="2",organization="CRL - Test",region="us-east4",store="2"} 0
    # HELP crdb_dedicated_addsstable_proposals Number of SSTable ingestions proposed (i.e. sent to Raft by lease holders)
    # TYPE crdb_dedicated_addsstable_proposals counter
    crdb_dedicated_addsstable_proposals{cluster="test-gcp",instance="172.28.0.167:8080",node="cockroachdb-j5t6j",node_id="1",organization="CRL - Test",region="us-east4",store="1"} 0
    crdb_dedicated_addsstable_proposals{cluster="test-gcp",instance="172.28.0.49:8080",node="cockroachdb-ttzj8",node_id="3",organization="CRL - Test",region="us-east4",store="3"} 0
    crdb_dedicated_addsstable_proposals{cluster="test-gcp",instance="172.28.0.99:8080",node="cockroachdb-r5rns",node_id="2",organization="CRL - Test",region="us-east4",store="2"} 0
    
  5. Once metrics export has been enabled and the scrape endpoint(s) tested, you need to configure your metrics collector to periodically poll the scrape endpoint(s). Configure your Prometheus configuration file's scrape_configs section as in the following example:

    icon/buttons/copy
    global:
      scrape_interval: 10s
      evaluation_interval: 10s
    
    # Prometheus configuration for CockroachDB Dedicated for a single region
    scrape_configs:
      - job_name: '{job_name}'
        metrics_path: '/api/v1/clusters/{cluster_id}/metricexport/prometheus/{cluster_region}/scrape'
        static_configs:
          - targets: ['cockroachlabs.cloud']
        scheme: 'https'
        authorization:
          credentials: '{secret_key}'
    

    Where:

    • {job_name} is a job name you assign to scraped metrics by default, such as scrape-cockroach-us-east4.
    • {cluster_id} is your CockroachDB Dedicated cluster ID as determined in step 1, resembling f78b7feb-b6cf-4396-9d7f-494982d7d81e.
    • {cluster_region} is a region of your CockroachDB Dedicated cluster as shown in the targets of step 3, such as us-east4. You can also find your cluster’s region(s) on the Cluster Overview page.
    • {secret_key} is your CockroachDB Dedicated API key. See API Access for instructions on generating this key.

Monitor the status of a metrics export configuration

To check the status of an existing AWS Cloudwatch metrics export configuration, use the following Cloud API command:

icon/buttons/copy
curl --request GET \
  --url https://cockroachlabs.cloud/api/v1/clusters/{cluster_id}/metricexport/cloudwatch \
  --header "Authorization: Bearer {secret_key}"

Where:

  • {cluster_id} is your CockroachDB Dedicated cluster's cluster ID, which can be found in the URL of your Cloud Console for the specific cluster you wish to configure, resembling f78b7feb-b6cf-4396-9d7f-494982d7d81e.
  • {secret_key} is your CockroachDB Dedicated API key. See API Access for instructions on generating this key.

To check the status of an existing Datadog metrics export configuration, use the following Cloud API command:

icon/buttons/copy
curl --request GET \
  --url https://cockroachlabs.cloud/api/v1/clusters/{cluster_id}/metricexport/datadog \
  --header "Authorization: Bearer {secret_key}"

Where:

  • {cluster_id} is your CockroachDB Dedicated cluster's cluster ID, which can be found in the URL of your Cloud Console for the specific cluster you wish to configure, resembling f78b7feb-b6cf-4396-9d7f-494982d7d81e.
  • {secret_key} is your CockroachDB Dedicated API key. See API Access for instructions on generating this key.

To check the status of an existing Prometheus metrics export configuration, use the following Cloud API command:

icon/buttons/copy
curl --request GET \
  --url https://cockroachlabs.cloud/api/v1/clusters/{cluster_id}/metricexport/prometheus \
  --header "Authorization: Bearer {secret_key}"

Where:

  • {cluster_id} is your CockroachDB Dedicated cluster's cluster ID, which can be found in the URL of your Cloud Console for the specific cluster you wish to configure, resembling f78b7feb-b6cf-4396-9d7f-494982d7d81e.
  • {secret_key} is your CockroachDB Dedicated API key. See API Access for instructions on generating this key.

Update an existing metrics export configuration

To update an existing CockroachDB Dedicated metrics export configuration, make any necessary changes to your cloud provider configuration (e.g., AWS CloudWatch, Datadog, or Prometheus), then issue the same POST Cloud API command as shown in the Enable metrics export instructions for your cloud provider with the desired updated configuration. Follow the Monitor the status of a metrics export configuration instructions to ensure the update completes successfully.

Disable metrics export

To disable an existing AWS Cloudwatch metrics export configuration, and stop sending metrics to Cloudwatch, use the following Cloud API command:

icon/buttons/copy
curl --request DELETE \
  --url https://cockroachlabs.cloud/api/v1/clusters/{cluster_id}/metricexport/cloudwatch \
  --header "Authorization: Bearer {secret_key}"

Where:

  • {cluster_id} is your CockroachDB Dedicated cluster's cluster ID, which can be found in the URL of your Cloud Console for the specific cluster you wish to configure, resembling f78b7feb-b6cf-4396-9d7f-494982d7d81e.
  • {secret_key} is your CockroachDB Dedicated API key. See API Access for instructions on generating this key.

To disable an existing Datadog metrics export configuration, and stop sending metrics to Datadog, use the following Cloud API command:

icon/buttons/copy
curl --request DELETE \
  --url https://cockroachlabs.cloud/api/v1/clusters/{cluster_id}/metricexport/datadog \
  --header "Authorization: Bearer {secret_key}"

Where:

  • {cluster_id} is your CockroachDB Dedicated cluster's cluster ID, which can be found in the URL of your Cloud Console for the specific cluster you wish to configure, resembling f78b7feb-b6cf-4396-9d7f-494982d7d81e.
  • {secret_key} is your CockroachDB Dedicated API key. See API Access for instructions on generating this key.

To disable an existing Prometheus metrics export configuration, and stop sending metrics to Prometheus, use the following Cloud API command:

icon/buttons/copy
curl --request DELETE \
  --url https://cockroachlabs.cloud/api/v1/clusters/{cluster_id}/metricexport/prometheus \
  --header "Authorization: Bearer {secret_key}"

Where:

  • {cluster_id} is your CockroachDB Dedicated cluster's cluster ID, which can be found in the URL of your Cloud Console for the specific cluster you wish to configure, resembling f78b7feb-b6cf-4396-9d7f-494982d7d81e.
  • {secret_key} is your CockroachDB Dedicated API key. See API Access for instructions on generating this key.

Limitations

  • Metrics export to AWS CloudWatch is only available on CockroachDB Dedicated clusters which are hosted on AWS. If your CockroachDB Dedicated cluster is hosted on GCP or Azure, you can export metrics to Datadog or Prometheus instead.
  • AWS CloudWatch does not currently support histograms. Any histogram-type metrics emitted from your CockroachDB Dedicated cluster are dropped by CloudWatch. See Prometheus metric type conversion for more information, and Logging dropped Prometheus metrics for instructions on tracking dropped histogram metrics in CloudWatch.

Troubleshooting

AWS CloudWatch

Be sure you are providing your own AWS Account ID as shown on the AWS IAM page to step 6, not the AWS cloud provider account ID as returned from step 3.

If you are using an existing AWS role, or are otherwise using a role name different from the example name used in this tutorial, be sure to use your own role name in step 8 in place of CockroachCloudMetricsExportRole.

Your CockroachDB Dedicated cluster must be running on AWS (not GCP or Azure) to make use of metrics export to AWS CloudWatch. If your CockroachDB Dedicated cluster is hosted on GCP or Azure, you can export metrics to Datadog or Prometheus instead.

See Also


Yes No
On this page

Yes No