A collection of Sensu Go checks for MySQL and MariaDB, converted from the sensu-plugins-mysql Ruby plugin collection.
| Binary | Description |
|---|---|
check-mysql-alive |
Verifies the server is reachable and accepting connections |
check-mysql-connections |
Monitors Threads_connected against warning/critical thresholds (absolute or percentage of max_connections) |
check-mysql-replication |
Checks replication IO/SQL thread status and lag; auto-detects MySQL vs MariaDB |
check-mysql-threads |
Monitors Threads_running with both high and low warning/critical thresholds |
check-mysql-disk |
Checks total database size from information_schema against thresholds in MB |
check-mysql-innodb-lock |
Detects InnoDB lock waits exceeding configurable time thresholds |
check-mysql-query-result-count |
Runs an arbitrary SQL query and alerts based on the row count or a scalar count value |
All checks share these connection flags:
| Flag | Short | Default | Description |
|---|---|---|---|
--hostname |
localhost |
MySQL host | |
--port |
3306 |
MySQL port | |
--socket |
-s |
Unix socket (overrides host/port) | |
--user |
-u |
MySQL user | |
--password |
-p |
MySQL password | |
--ini |
-i |
Path to my.cnf file; credentials from the ini file override --user/--password |
|
--ini-section |
client |
Section to read from the ini file | |
--database |
-d |
test |
Database to connect to |
Connects to MySQL and returns OK if the ping succeeds.
check-mysql-alive --hostname db1.example.com --user sensu --password secret
Alerts when the number of connected threads exceeds thresholds. Use --percentage to compare against max_connections instead of an absolute count.
check-mysql-connections --warning 100 --critical 128
check-mysql-connections --warning 80 --critical 90 --percentage
| Flag | Short | Default | Description |
|---|---|---|---|
--warning |
-w |
100 |
Warning threshold |
--critical |
-c |
128 |
Critical threshold |
--percentage |
false |
Use percentage of max_connections |
Checks that the IO and SQL replication threads are running and that replication lag is within thresholds. Automatically uses the correct queries for MySQL (via performance_schema) and MariaDB (via SHOW REPLICA STATUS).
check-mysql-replication --warning 900 --critical 1800
| Flag | Short | Default | Description |
|---|---|---|---|
--warning |
-w |
900 |
Warning threshold for replication lag (seconds) |
--critical |
-c |
1800 |
Critical threshold for replication lag (seconds) |
--master-connection |
-m |
Replication master connection name |
Monitors Threads_running. Supports both high thresholds (too many active threads) and low thresholds (useful for detecting silent failures on replicas).
check-mysql-threads --warning 50 --critical 100
check-mysql-threads --warning 50 --critical 100 --warning-low 2 --critical-low 1
| Flag | Short | Default | Description |
|---|---|---|---|
--warning |
-w |
50 |
High warning threshold |
--critical |
-c |
100 |
High critical threshold |
--warning-low |
0 |
Low warning threshold (0 = disabled) | |
--critical-low |
0 |
Low critical threshold (0 = disabled) |
Queries information_schema.tables for total database size and alerts when it exceeds thresholds (in MB). Use --schema to check a single database; omit to check all schemas.
check-mysql-disk --warning 5120 --critical 10240
check-mysql-disk --schema myapp --warning 1024 --critical 2048
| Flag | Short | Default | Description |
|---|---|---|---|
--warning |
-w |
5120 |
Warning threshold in MB |
--critical |
-c |
10240 |
Critical threshold in MB |
--schema |
-S |
Schema to measure (empty = all schemas) |
Detects InnoDB lock waits that have been blocking for longer than the warning threshold. Automatically uses performance_schema.data_lock_waits on MySQL 8.0+ and information_schema.INNODB_LOCK_WAITS on MySQL 5.x and MariaDB.
check-mysql-innodb-lock --warning 5 --critical 10
| Flag | Short | Default | Description |
|---|---|---|---|
--warning |
-w |
5 |
Seconds a lock wait must exceed to be reported |
--critical |
-c |
10 |
Seconds a lock wait must exceed to trigger critical |
Runs an arbitrary SQL query and compares the result count against thresholds. By default counts the number of rows returned. Use --count-value for SELECT COUNT(*) queries to read the scalar value instead.
# Alert when more than 10 rows match a condition
check-mysql-query-result-count --query "SELECT id FROM orders WHERE status='failed'" \
--warning 10 --critical 50
# Alert when a COUNT(*) query exceeds thresholds
check-mysql-query-result-count --query "SELECT COUNT(*) FROM jobs WHERE created_at < NOW() - INTERVAL 1 HOUR" \
--count-value --warning 100 --critical 500
| Flag | Short | Default | Description |
|---|---|---|---|
--query |
-q |
SQL query to execute (required) | |
--warning |
-w |
Warning threshold (required if --critical not set) |
|
--critical |
-c |
Critical threshold (required if --warning not set) |
|
--count-value |
false |
Read the integer value from the first column of the first row instead of counting rows |