Cluster Management
Clusters provide a way to group related hosts for streamlined management and batch operations. This guide covers cluster creation, management, and common usage patterns.
Creating Clusters
Section titled “Creating Clusters”Navigate to Host Management → Clusters → Create Cluster:
- Name — A unique, descriptive name (e.g.,
prod-web-servers) - Description — Optional notes about the cluster’s purpose
- Host Selection — Add hosts manually or by tag filter
Tag-based Dynamic Groups
Section titled “Tag-based Dynamic Groups”Instead of manually adding hosts, you can use tag filters to create dynamic groups:
env:production AND role:webHosts matching the filter are automatically included. When new hosts are tagged, they join the cluster automatically.
Cluster Operations
Section titled “Cluster Operations”Once a cluster is defined, you can perform operations on all its hosts simultaneously:
- Batch Script Execution — Run any script across all cluster hosts
- Rolling Updates — Execute scripts with configurable concurrency and delay between hosts
- Health Overview — See aggregated health metrics for the entire cluster
- Bulk Tag Management — Apply or remove tags from all hosts in the cluster
Common Cluster Patterns
Section titled “Common Cluster Patterns”By Environment
Section titled “By Environment”production → all hosts tagged env:productionstaging → all hosts tagged env:stagingdevelopment → all hosts tagged env:developmentBy Service Role
Section titled “By Service Role”web-servers → role:webdb-servers → role:databaseworkers → role:workercache → role:cacheBy Region
Section titled “By Region”us-east → region:us-east-1eu-west → region:eu-west-1Cluster Dashboard
Section titled “Cluster Dashboard”Each cluster has a dashboard showing:
| Metric | Description |
|---|---|
| Total Hosts | Number of hosts in the cluster |
| Online | Hosts with active agent connections |
| Avg CPU | Average CPU utilization across the cluster |
| Avg Memory | Average memory usage |
| Recent Tasks | Last 10 batch executions |
Best Practices
Section titled “Best Practices”- Use consistent tagging conventions across your infrastructure
- Prefer tag-based dynamic groups over static host lists for production clusters
- Set up cluster-specific alerting thresholds
- Document cluster purposes in the description field
- Review cluster membership regularly to remove decommissioned hosts