clickhouse-zabbix

Template to monitor main Clickhouse parameters without using additional scripts

It uses HTTP interface to read metrics from system database using Zabbix Agent's HTTP client

Gathered metrics:

• MaxPartCountForPartition
• ReplicasMaxQueueSize
• Uptime
• InsertedBytes
• InsertedRows
• Query
• InsertQuery
• SelectQuery
• Merge
• Inactive parts count
• HTTPConnection
• TCPConnection
• InterserverConnection
• LeaderReplica
• ReadonlyReplica
• Used memory (this is roughly calculated from metrics and is not accurate)
• Ping (/ping HTTP endpoint)

Requirements:

• Zabbix 4.2+ (for Javascript support)
• Clickhouse 19.16+ (tested on, but should work on earlier versions)

Macros:

• {$CH_URL} Clickhouse URL (http://127.0.0.1:8123 default)
• {$CH_MAX_INACTIVE_PARTS} for triggering on too much inactive parts
• {$CH_MAX_PARTS} for triggering on too much parts in partition
• {$CH_MAX_REPLICA_QUEUE} for triggering on too large replication queue