GCP Dataflow¶
Collect GCP Dataflow metrics.
Configuration¶
Install Func¶
We recommend enabling TrueWatch Integration - Extensions - Managed Func. All prerequisites are installed automatically; continue with the script installation.
If you deploy Func yourself, see Self-deploying Func.
Install Script¶
- Log in to the Func console, click [Script Market], open the TrueWatch Script Market, and search for
integration_gcp_dataflow. - Click [Install], then enter the required parameters: account file name, Target Principal, and account name.
- Click [Deploy Startup Script]. The system automatically creates a
Startupscript set and configures the corresponding startup script. - After enabling it, find the automatic trigger configuration under "Management / Automatic Trigger Configuration". Click [Execute] to run it immediately, then review the task record and logs.
Verification¶
- Under "Management / Automatic Trigger Configuration", confirm that the task exists and check its task records and logs for errors.
- In TrueWatch, check "Infrastructure / Custom" for the corresponding asset information.
- In TrueWatch, check "Metrics" for the corresponding measurement and monitoring data.
Metrics¶
GCP Dataflow metrics are stored in the gcp_dataflow measurement. The original metrics come from the Cloud Monitoring dataflow.googleapis.com/ namespace. The table below lists key metrics; see GCP Dataflow metrics for the complete metric catalog.
| Metric | Original Metric | Description | Unit |
|---|---|---|---|
| job_active_worker_instances | job/active_worker_instances |
Job active worker instances | Count |
| job_aggregated_worker_utilization | job/aggregated_worker_utilization |
Job aggregated worker utilization | % |
| job_backlog_bytes | job/backlog_bytes |
Job backlog bytes | Bytes |
| job_backlog_elements | job/backlog_elements |
Job backlog elements | Count |
| job_current_num_vcpus | job/current_num_vcpus |
Job current number vcpus | Count |
| job_current_shuffle_slots | job/current_shuffle_slots |
Job current shuffle slots | Count |
| job_data_watermark_age | job/data_watermark_age |
Job data watermark age | s |
| job_estimated_bytes_consumed_count | job/estimated_bytes_consumed_count |
Job estimated bytes consumed count | Bytes |
| job_estimated_bytes_produced_count | job/estimated_bytes_produced_count |
Job estimated bytes produced count | Bytes |
| job_is_bottleneck | job/is_bottleneck |
Job is bottleneck | Bool |
| job_is_failed | job/is_failed |
Job is failed | Count |
| job_memory_capacity | job/memory_capacity |
Job memory capacity | Bytes |
| job_system_lag | job/system_lag |
Job system lag | s |
| job_target_worker_instances | job/target_worker_instances |
Job target worker instances | Count |
| job_total_memory_usage_time | job/total_memory_usage_time |
Job total memory usage time | GiBy.s |