| name | managing-clusters |
| description | Monitor and manage cluster health, nodes, tasks, and settings. Use when checking cluster status, troubleshooting node issues, investigating shard allocation problems, or managing cluster settings.
|
Cluster Management
Health & Status
| Function | Example |
|---|
ES_CLUSTER_HEALTH(index?) | ES_CLUSTER_HEALTH() |
ES_CLUSTER_STATE(metrics) | ES_CLUSTER_STATE('nodes,routing_table') |
ES_CLUSTER_STATS() | ES_CLUSTER_STATS() |
ES_CLUSTER_INFO() | ES_CLUSTER_INFO() |
Node Management
| Function | Example |
|---|
ES_NODES_INFO(node_id?, metrics?) | ES_NODES_INFO('_all', 'os,jvm') |
ES_NODES_STATS(node_id?, metrics?) | ES_NODES_STATS('_all', 'fs,jvm') |
ES_NODE_HOT_THREADS(node_id?) | ES_NODE_HOT_THREADS('_all') |
Task Management
| Function | Example |
|---|
ES_LIST_TASKS(detailed?) | ES_LIST_TASKS(TRUE) |
ES_PENDING_TASKS() | ES_PENDING_TASKS() |
ES_CANCEL_TASK(task_id) | ES_CANCEL_TASK('node1:12345') |
Shard Allocation
| Function | Example |
|---|
ES_ALLOCATION_EXPLAIN(index?, shard?, primary?) | Debug allocation |
Settings
| Function | Example |
|---|
ES_CLUSTER_SETTINGS(include_defaults?) | ES_CLUSTER_SETTINGS(TRUE) |
ES_UPDATE_CLUSTER_SETTINGS(settings) | Update settings |
Health Check Pattern
DECLARE health DOCUMENT;
SET health = ES_CLUSTER_HEALTH();
PRINT 'Status: ' || health['status'];
PRINT 'Nodes: ' || health['number_of_nodes'];
PRINT 'Unassigned: ' || health['unassigned_shards'];
IF health['status'] != 'green' THEN
DECLARE explain DOCUMENT;
SET explain = ES_ALLOCATION_EXPLAIN();
PRINT 'Issue: ' || explain['explanation'];
END IF;
Pre-built Skills
| Skill | Description |
|---|
RUN SKILL cluster_health_check(index?) | Full health report |
RUN SKILL node_stats_summary() | Resource overview |
RUN SKILL explain_allocation() | Debug shards |
Troubleshooting
Red cluster: Check ES_CLUSTER_HEALTH() → ES_ALLOCATION_EXPLAIN() → check disk with ES_NODES_STATS('_all', 'fs')
Yellow cluster: Usually replica allocation - check node count vs replica settings
High heap: Check ES_NODES_STATS('_all', 'jvm') for nodes with heap_used_percent > 85%
