Skip to main content

Monitoring Metrics for Warehouse CN Groups

StarRocks provides a variety of metrics for monitoring and managing Compute Node Groups (CN Groups) in warehouses.

Metric items​

warehouse_cngroup​

  • Type: Gauge/Counter
  • Description: Warehouse Compute Node Group metrics with different field labels to monitor various aspects of CN Group performance and health.

Field labels​

cngroup_nodes_count​
  • Type: Gauge
  • Description: Total number of compute nodes in the CN Group.
cngroup_alive_nodes_count​
  • Type: Gauge
  • Description: Number of compute nodes alive in the CN Group.
running_queries_count​
  • Type: Gauge
  • Description: Number of queries currently running in the CN Group on the current Frontend (FE).
cngroup_status​
  • Type: Gauge
  • Description: Status of the CN Group. Valid values: 0 (disabled) and 1 (enabled).
scheduled_queries_count​
  • Type: Counter
  • Description: Total number of queries scheduled to the CN Group.
success_queries_count​
  • Type: Counter
  • Description: Total number of successfully executed queries in the CN Group.
failed_queries_count​
  • Type: Counter
  • Description: Total number of failed queries in the CN Group.
query_max_latency_ms​
  • Type: Gauge
  • Description: Maximum query latency (in milliseconds) for the CN Group.
query_avg_latency_ms​
  • Type: Gauge
  • Description: Average query latency (in milliseconds) for the CN Group.
avg_cpu_used_permille​
  • Type: Gauge
  • Description: Average CPU usage in permille (per thousand) across all compute nodes in the CN Group. -1.0 is returned if the value is invalid or unavailable.
max_compute_node_running_queries_count​
  • Type: Gauge
  • Description: Maximum number of running queries across all compute nodes in the CN Group. -1 is returned if the value is invalid or unavailable.

Usage examples​

Monitor CN Group health and performance​

You can use these metrics to monitor the health and performance of your warehouse CN Groups:

# Check the availability of nodes in the CN Group 
warehouse_cngroup{field="cngroup_alive_nodes_count"} / warehouse_cngroup{field="cngroup_nodes_count"}

# Monitor CN Group status
warehouse_cngroup{field="cngroup_status"}

# Check query success rate
warehouse_cngroup{field="success_queries_count"} / warehouse_cngroup{field="scheduled_queries_count"}

# Monitor query latency
warehouse_cngroup{field="query_avg_latency_ms"}

# Check CPU utilization
warehouse_cngroup{field="avg_cpu_used_permille"} / 10

Metric labels​

All warehouse CN Group metrics include the following labels:

  • warehouse_id: The unique identifier of the warehouse
  • warehouse_name: The name of the warehouse
  • cngroup_name: The name of the CN Group
  • field: The specific field being measured (as listed above)

These labels allow you to filter and group metrics by specific warehouses and CN Groups to monitor their individual performance characteristics.

Performance considerations​

  • The CN Group resource usage metrics are cached for 1 second to avoid excessive computation
  • CPU usage metrics return -1.0 when the value is invalid, null, or NaN
  • Maximum running queries count returns -1 when the value is invalid or unavailable
  • Query latency metrics are updated atomically to ensure thread safety

Metric examples​

Example 1: Warehouse CN Group metrics demo:

{"tags":{"metric":"warehouse_cngroup","field":"cngroup_nodes_count","warehouse_id":"0","warehouse_name":"default_warehouse","cngroup_name":"_builtin_cngroup_0_"},"unit":"nounit","value":2},
{"tags":{"metric":"warehouse_cngroup","field":"cngroup_alive_nodes_count","warehouse_id":"0","warehouse_name":"default_warehouse","cngroup_name":"_builtin_cngroup_0_"},"unit":"nounit","value":2},
{"tags":{"metric":"warehouse_cngroup","field":"running_queries_count","warehouse_id":"0","warehouse_name":"default_warehouse","cngroup_name":"_builtin_cngroup_0_"},"unit":"nounit","value":0},
{"tags":{"metric":"warehouse_cngroup","field":"cngroup_status","warehouse_id":"0","warehouse_name":"default_warehouse","cngroup_name":"_builtin_cngroup_0_"},"unit":"nounit","value":1},
{"tags":{"metric":"warehouse_cngroup","field":"scheduled_queries_count","warehouse_id":"0","warehouse_name":"default_warehouse","cngroup_name":"_builtin_cngroup_0_"},"unit":"nounit","value":98},
{"tags":{"metric":"warehouse_cngroup","field":"success_queries_count","warehouse_id":"0","warehouse_name":"default_warehouse","cngroup_name":"_builtin_cngroup_0_"},"unit":"nounit","value":83},
{"tags":{"metric":"warehouse_cngroup","field":"failed_queries_count","warehouse_id":"0","warehouse_name":"default_warehouse","cngroup_name":"_builtin_cngroup_0_"},"unit":"nounit","value":15},
{"tags":{"metric":"warehouse_cngroup","field":"query_max_latency_ms","warehouse_id":"0","warehouse_name":"default_warehouse","cngroup_name":"_builtin_cngroup_0_"},"unit":"nounit","value":1485.0},
{"tags":{"metric":"warehouse_cngroup","field":"query_avg_latency_ms","warehouse_id":"0","warehouse_name":"default_warehouse","cngroup_name":"_builtin_cngroup_0_"},"unit":"nounit","value":54.255102040816325},
{"tags":{"metric":"warehouse_cngroup","field":"avg_cpu_used_permille","warehouse_id":"0","warehouse_name":"default_warehouse","cngroup_name":"_builtin_cngroup_0_"},"unit":"nounit","value":54.255102040816325},
{"tags":{"metric":"warehouse_cngroup","field":"max_compute_node_running_queries_count","warehouse_id":"0","warehouse_name":"default_warehouse","cngroup_name":"_builtin_cngroup_0_"},"unit":"nounit","value":0},