Integrate DataDog via Terraform

DataDog is a monitoring and security platform for cloud-scale applications. It provides real-time visibility into servers, databases, applications, and services. You can use DataDog to track the performance of your CelerData Cloud BYOC clusters, identify bottlenecks, and ensure optimal operation, thereby maintaining the health and efficiency of the cluster.

CelerData supports integrating DataDog with your CelerData cluster by installing the DataDog Agent via Terraform. You can follow these steps to install the DataDog Agent:

Step 1: Prepare and upload script

To install the DataDog Agent for a CelerData cluster, you must first prepare the script file that installs the Agent, and upload the script to the S3 bucket on which all EC2 instances that host the CelerData cluster have the required read or write permissions. For more information on running scripts in CelerData, see Run Scripts.

1. Prepare the DataDog Agent installation script install_datadog.sh:

   on_deployment_completed() {
       # Replace <YOUR_DD_KEY> with your DataDog API key.
       # Replace <YOUR_DD_SITE> with your DataDog site URL.
       DD_API_KEY=<YOUR_DD_API_KEY> DD_SITE="<YOUR_DD_SITE>" bash -c "$(curl -L https://s3.amazonaws.com/dd-agent/scripts/install_script_agent7.sh)"
   
       if [ "$STARROCKS_INSTANCE_TYPE" = "FE" ]; then
           fe_config='
   init_config:
   instances:
       - openmetrics_endpoint: http://127.0.0.1:8030/metrics
       # Repleace <fe_metric_namespace> with the namespace you want to use for FE/Coordinator Node metrics.
       namespace: <fe_metric_namespace>
       metrics:
           - starrocks.*
   '
       sudo bash -c "echo '$fe_config' > /etc/datadog-agent/conf.d/openmetrics.d/conf.yaml"
       elif [ "$STARROCKS_INSTANCE_TYPE" = "BE" ]; then
           be_config='
   init_config:
   instances:
       - openmetrics_endpoint: http://127.0.0.1:8040/metrics
       # Repleace <be_metric_namespace> with the namespace you want to use for BE/Compute Node metrics.
       namespace: <be_metric_namespace>
       metrics:
           - starrocks.*
   '
           sudo bash -c "echo '$be_config' > /etc/datadog-agent/conf.d/openmetrics.d/conf.yaml"
       else
           echo "UNKNOWN INSTANCE TYPE: $STARROCKS_INSTANCE_TYPE"
           exit 1
       fi
   
       sudo systemctl stop datadog-agent
       sudo systemctl start datadog-agent
   }

   You must specify the following parameter in the script:

   Parameter	Description
   YOUR_DD_API_KEY	Your DataDog API key.
   YOUR_DD_SITE	Your DataDog site URL.
   fe_metric_namespace	The namespace for FE/Coordinator Node metrics. It is recommended to set this parameter in the format cluster_name + xxx + fe to allow you to identify the metrics easily. Example: celerdata_test_fe.
   be_metric_namespace	The namespace for BE/Compute Node metrics. It is recommended to set this parameter in the format cluster_name + xxx + fe to allow you to identify the metrics easily. Example: celerdata_test_be.

2. Upload install_datadog.sh to the S3 bucket whose read permissions were granted to the EC2 instances that host your CelerData cluster.

Step 2: Install Agent via Terraform

After uploading the file(s) to S3, you can install the DataDog Agent by running the script while deploying the cluster.

1. Declare install_datadog.sh in the celerdatabyoc_classic_cluster or celerdatabyoc_elastic_cluster resource used to deploy your CelerData cluster.

   The following example shows only the script-related arguments in a celerdatabyoc_classic_cluster resource. For detailed information on other arguments, see celerdatabyoc_classic_cluster and celerdatabyoc_elastic_cluster.

   resource "celerdatabyoc_classic_cluster" "cluster-with-datadog" {
     ...
     init_scripts {
       // Replace <path_to_script_logs> with a directory for the script log storage.
       logs_dir    = "<path_to_script_logs>"
       // Replace <path_to_install_datadog.sh> with the S3 URI of install_datadog.sh.
       script_path = "<path_to_install_datadog.sh>"
     }
     
     run_scripts_parallel = false
     ...
   }

   You must specify the following argument in the resource:

   Parameter	Description
   path_to_script_logs	The S3 path used to store the logs of your scripts. The S3 path is in one of the following formats: s3://<bucket_name>/<folder_name>/: You must go to the instance profiles associated with the EC2 instances to which the CelerData cluster will be deployed and configure read and write permissions on the specified bucket. s3a://<bucket_name>/<folder_name>/: You must go to the instance profiles associated with the EC2 instances to which the CelerData cluster will be deployed and configure read and write permissions on the specified bucket. /<folder_name>/: You can use this format if you want the logs to be stored in the bucket you use for the cluster deployment. Tip It is recommended to specify this argument as the parent directory of install_datadog.sh.
   path_to_install_datadog.sh	The S3 URI of install_datadog.sh. The S3 URI is in one of the following formats: s3://<bucket_name>/<folder_name>/install_datadog.sh: You must go to the instance profiles associated with the EC2 instances to which the CelerData cluster will be deployed and configure read permissions on the specified bucket. s3a://<bucket_name>/<folder_name>/install_datadog.sh: You must go to the instance profiles associated with the EC2 instances to which the CelerData cluster will be deployed and configure read permissions on the specified bucket. /<folder_name>/install_datadog.sh: You can use this format if install_datadog.sh is stored in the bucket you use for the cluster deployment.

2. Apply the resource.

   terraform apply

If you have installed the DataDog Agent successfully, you will find the metrics of your CelerData cluster on the Metrics Explorer page of your DataDog site. For the detailed descriptions of the metrics supported, refer to StarRocks Monitoring Metrics.