MariaDB Integration
Monitor query performance, thread pool utilization, and Galera cluster replication health across your MariaDB deployments. Get AI-powered anomaly detection and predictive alerts before incidents impact production.
How It Works
Install the TigerOps Agent
Deploy the TigerOps agent on your MariaDB host or as a sidecar. The agent auto-discovers MariaDB via the performance_schema and information_schema, requiring only a read-only monitoring user.
Configure Database Credentials
Create a dedicated monitoring user with SELECT privileges on performance_schema. Set the credentials in the agent config — no superuser access is needed for full metric coverage.
Enable Galera Cluster Metrics
For Galera clusters, the agent automatically collects wsrep_* status variables — cluster size, replication lag, flow control, and node state — across all cluster members.
Set Alert Thresholds
Define SLOs for query latency, replication lag, and connection pool saturation. TigerOps fires predictive alerts and correlates Galera replication stalls with upstream write pressure.
What You Get Out of the Box
Query Performance Analysis
Top slow queries ranked by total execution time, lock wait time, and rows examined. TigerOps parses the performance_schema digest table to surface normalized query patterns without exposing sensitive data.
Thread Pool Monitoring
Track MariaDB thread pool queue depth, active threads, waiting threads, and connection concurrency. Alerts fire when pool saturation predicts connection refusals.
Galera Cluster Replication Health
Monitor wsrep_local_recv_queue, wsrep_flow_control_paused, cluster size, and node state transitions. TigerOps detects split-brain risks and replication stalls in real time.
InnoDB Buffer Pool Efficiency
Track buffer pool hit ratio, dirty page percentage, read-ahead efficiency, and checkpoint age. Correlate buffer pool pressure with query latency spikes automatically.
Replication Lag Tracking
Per-replica seconds_behind_master with predictive alerting. TigerOps monitors semi-sync replication acknowledgment times and binlog position drift across all replicas.
AI Root Cause Analysis
When query latency degrades, TigerOps AI correlates InnoDB lock wait events, Galera flow control pauses, and thread pool starvation to surface the root cause instantly.
TigerOps Agent Config for MariaDB
Add this block to your TigerOps agent configuration file to start collecting MariaDB metrics.
# TigerOps MariaDB integration config
# Place at /etc/tigerops/conf.d/mariadb.yaml
integrations:
- name: mariadb
type: mariadb
config:
host: 127.0.0.1
port: 3306
user: tigerops_monitor
# Store password in environment or secrets manager
password: "${MARIADB_MONITOR_PASSWORD}"
tls:
enabled: true
ca_cert: /etc/ssl/mariadb/ca.pem
# Galera cluster monitoring (set to false for standalone)
galera:
enabled: true
collect_wsrep_status: true
# Query performance schema
performance_schema:
enabled: true
digest_sample_rate: 1.0 # fraction of digests to collect
top_query_count: 20 # top N queries by total time
# Replication monitoring
replication:
enabled: true
collect_replica_status: true
scrape_interval: 15s
remote_write:
endpoint: https://ingest.atatus.net/api/v1/write
bearer_token: "${TIGEROPS_API_KEY}"
# Monitoring user setup (run once):
# CREATE USER 'tigerops_monitor'@'127.0.0.1'
# IDENTIFIED BY 'secure_password';
# GRANT SELECT ON performance_schema.* TO 'tigerops_monitor'@'127.0.0.1';
# GRANT SELECT ON information_schema.* TO 'tigerops_monitor'@'127.0.0.1';
# GRANT PROCESS, REPLICATION CLIENT ON *.* TO 'tigerops_monitor'@'127.0.0.1';
# FLUSH PRIVILEGES;Common Questions
Which MariaDB versions does TigerOps support?
TigerOps supports MariaDB 10.4 and later, including MariaDB 10.5, 10.6, 10.11 LTS, and 11.x. Both standalone instances and Galera clusters are supported. Amazon RDS for MariaDB and Azure Database for MariaDB are also supported via their respective metric endpoints.
What privileges does the monitoring user need?
The monitoring user requires SELECT on performance_schema.*, SELECT on information_schema.*, and PROCESS and REPLICATION CLIENT global privileges. No DML privileges are needed. TigerOps provides a ready-made SQL snippet to create the monitoring user with the exact required permissions.
How does TigerOps monitor Galera clusters?
TigerOps connects to each Galera node independently and collects wsrep status variables. It computes cluster-wide metrics such as effective replication lag, flow control ratio, and cluster size delta. Alerts are de-duplicated across nodes so you get one incident per cluster event.
Can TigerOps detect deadlocks and lock contention?
Yes. TigerOps monitors innodb_deadlocks, innodb_row_lock_waits, and innodb_row_lock_time from information_schema.INNODB_STATUS. It also parses performance_schema events_waits_summary to pinpoint which query patterns are causing the most lock contention.
Does TigerOps support MariaDB ColumnStore?
Yes. TigerOps collects ColumnStore-specific metrics including InfoManager queue depth, DML write rates, and extent map utilization alongside standard InnoDB metrics, giving you a unified view across mixed storage engines.
Stop Flying Blind on MariaDB Performance
Query performance analysis, Galera cluster health, and AI root cause analysis. Deploy in 5 minutes.